Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Dear friends and participants of the 2009 MR Angio Club Meeting It is with great pleasure that we welcome everyone back to Michigan State University for the 21st International Conference in MR Angiography. Since being founded by a team lead by Dr. Jim Potchen in 1989, Michigan State University has been proud to host the annual meeting on three previous occasions, the last in 1994. The MR Angio Club annual meeting has spanned the globe from Asia, to Europe, to the USA and Canada as well as the Far East. This wide diversity in locality has been matched with an ever-growing diversity and ingenuity to produce ever-improving clinical and research applications of MR Angiography. The Local Organizing Team is excited about the excellent quality of the greater than 115 abstracts submitted to the 21st MR Angiography meeting. Both the oral presentations and the expanded poster section should provide very stimulating discussions typical of the MRA Club meetings. We are also proud to announce a Student Poster Award as part of this year’s meeting. MSU Kellogg Center will be an excellent facility for our scientific program with a large, comfortable stadium seating auditorium and hotel rooms to accommodate many of the presenters in one building. The Local Organizing Team has also arranged an exciting social evening program to highlight the Michigan State University campus and the history of Michigan that we hope you will enjoy. We are grateful for the ongoing support of our corporate sponsors who continue to support such stimulating meetings. The sponsors are listed elsewhere in this program an on the meeting homepage. We are also indebted to the London, Ontario Team including Janette Wallace, Johanne Guillemette, and Kellie Griffin. We hope you enjoy the 21st International Conference on MR Angiography and welcome back to Michigan State University. Local Organizing Team Jim Siebert – Scientific Committee Chairman Arlene Sierra – MRA Club International Secretary Stella Cash – MSU Alumni Interim Executive Director Tom Cooper – Vice Chairman MSU Radiology Colleen Hammond – Chief MSU MR Technologist Kevin DeMarco – President MRA Club 2009 MR-Angioclub East Lansing 2009 1 Conference Organization President President Elect J. Kevin DeMarco Martin Prince Past Presidents Executive Board James Potchen Georg Bongartz Dennis L. Parker Georg Bongartz Kevin DeMarco James Potchen James E. Siebert Phillipe Douek Martin Prince Paolo Pavone Charles Dumoulin Stephen J. Riederer Dennis L. Parker Christoph Herborn Brian Rutt Jorg F. Debatin John Huston Kazuhiko (Kaz)Sadamoto John Huston Frank R. Korosec David Saloner E. Kent Yucel Gerhard Laub Klaus Scheffler Frank R. Korosec Debiao Li Stefan Schoenberg David Saloner James F.M. Meaney James E. Siebert James F. M. Meaney Local Organizing Committee Stella Cash Tom Cooper Colleen Hammond Kevin Henley Cheryl Hunley Arlene Sierra Graham Wright Hitoshi Miki Brian K. Rutt Charles Mistretta Debiao Li Stephen J. Riederer Program Committee Congress Organizing Team James Siebert James Siebert Conference Coordinator Kellie Griffin Johanne Guillemette Carolyn Pakula Terri Schneider Registration Service www.registrationassistant.com 2 Executive Board Secretariat Arlene Sierra Congress Office @ Robarts Research Institute Johanne Guillemette Janette Wallace Conference Manager Secretariat Janette Wallace Accommodation www.hfs.msu.edu/kellogg MR-Angioclub East Lansing 2009 The 2009 MR Angio Conference Timetable Tuesday, September 29, 2009 07:30 – 9:00 Registration 09:00 – 9:30 Opening Ceremony 09:30 – 10:30 Selected Non- and Low-Gd Contrast Methods and Applications 10:30 – 11:00 Coffee Break 11:00 – 12:00 Compressed Sensing and HYPR 12:00 – 01:00 Lunch 01:00 – 02:45 Plaque Characterization and Vessel Wall Imaging 02:45 – 03:15 Coffee Break 03:15 – 05:00 Peripheral MRA Wednesday, September 30, 2009 08:00 – 09:45 Contrast Agents Performance and Safety 09:45 – 10:15 Coffee Break 10:15 – 12:00 Hemodynamics and Perfusion 12:00 – 01:00 Lunch 01:00 – 02:45 Head and Neck MRA 02:45 – 03:15 Coffee Break 03:15 – 05:00 Thoracic MRA Thursday, October 1, 2009 08:30 – 09:45 Cardiac Imaging / Coronary MRA 09:45 – 10:15 Coffee Break 10:15 – 12:00 Abdominal MRA 12:00 – 01:00 Lunch 01:00 – 02:45 Venous Imaging and Beyond 02:45 – 03:15 Coffee Break 03:15 – 05:00 New Horizons and Challenges in MRA MR-Angioclub East Lansing 2009 3 MR Angio 2009 Club appreciates the generous support of the following sponsors: PLATINUM SPONSORS GOLD SPONSORS SILVER SPONSORS 4 MR-Angioclub East Lansing 2009 MR-Angioclub East Lansing 2009 5 1. Registration, Internet Access 2. Speaker Ready Room 3. Main Meeting Room 4. Lunch & Breakfast 5. Exhibits & Posters, Coffee Breaks 6. Board Meeting Floor Plan for Exhibitors and Posters - Big 10 Room B (see pages 23-24) 6 MR-Angioclub East Lansing 2009 General Information Venue Kellogg Hotel and Conference Center Michigan State University East Lansing, Michigan Registration Foyer of Kellogg Hotel and Conference Center, Lobby Level, across from Auditorium. Message Board There will be a message board near the registration desk. Coffee Breaks & Lunches Snacks and drinks will be provided on the Lobby Level of the Kellogg Hotel and Conference Center in the Big Ten Room C. Internet Access Wireless Internet is available throughout the conference center. Speakers Speakers are requested to contact the speaker ready room (101, lobby level, right beside the Auditorium) and to hand in their Power Point presentation (on CD-ROM or data stick) at least 90 minutes before the start of the session of their presentation. Trained staff will be available to assist you with equipment. You can retrieve your CD-ROM or USB stick once your presentation has been uploaded. Your presentation will be deleted from the server after your talk and thus not made accessible to third parties. Posters Posters have to be mounted at the assigned stands in the poster Exhibition Area (Big Ten Room B) the morning of Tuesday September 29. Please dismount the poster Thursday October 1 by 5 pm. Remaining posters will be discarded. MR-Angioclub East Lansing 2009 7 Spousal Program Tuesday, September 29, 2009 09:00 – 03:30 pm Bus leaves Kellogg Center at 09:00 am Tour of the Gerald R. Ford Presidential Museum Lunch at Amway Grand Meijer Sculpture Gardens Bus leaves at 03:30 pm from Meijer Gardens back to the Kellogg Center Wednesday, September 30, 2009 08:30 – 03:30 pm Bus leaves the Kellogg Center at 08:30 am Trip to Frankenmuth Lunch at “Zehnders” Restaurant Bus leaves at 03:00 pm from Zehnders back to the Kellogg Center Thursday, October 1, 2009 Tour of Michigan State University Campus Lunch is provided 8 MR-Angioclub East Lansing 2009 Social Events Monday, September 28, 2009 07:00 – 09:30 pm Welcome Reception Radiology Department of Michigan State University Tuesday, September 29, 2009 07:00 – 10:00 pm Dinner at the University Club of Michigan State University Wednesday, September 30, 2009 07:00 – 10:00 pm Evening reception at the State Capital Michigan Historical Museum Dinner catered by Morton’s Thursday, October 1, 2009 07:00 – 10:00 pm Casual dining in a lakeside setting with Motown music. MR-Angioclub East Lansing 2009 9 Tuesday, September 29, 2009 9:00 am – 9:30 am Opening Ceremony 9:00 – 9:05 J. Kevin DeMarco MR Angio Club President 9:05 – 9:15 Lou Anna K. Simon President, Michigan State University 9:15 – 9:20 J. Ian Gray Vice President for Research and Graduate Studies 9:20 – 9:25 James Randolph Hillard Associate Provost for Human Health Affairs 9:25 – 9:30 James E. Siebert Program Director 9:30 am – 10:30 am Session 1 Selected Non - and Low- Gd Contrast Methods and Applications Session Chairs: Stephen J. Riederer, Manuela Aschauer 9:30 Liesbeth Geerts 1.1 Non-CE Imaging of the Pulmonary Arteries 9:42 Kevin Johnson 1.2 Accelerated Time Resolved Inflow with 3D Radial bSSFP 9:54 Gerhard Laub 1.3 Low - Dose 4D MR Angiography 10:06 Samuel Fielden 1.4 Balanced-gradient TSE for Non-contrast Peripheral MRA 10:18 Jing Liu 1.5 Self-gated Free Breathing 3D Cardiac Cine Imaging with Data Acquisition During Slice Encoding 10:30 am – 11:00 am 10 Coffee Break MR-Angioclub East Lansing 2009 11:00 am – 12:00 pm Session 2 Compressed Sensing and HYPR Session Chairs: Mark A Griswold, Dennis L. Parker 11:00 Mark Griswold 2.1 A Simple View of Compressed Sensing and How it Could Change Everything We Do in MRI and MRA 11:12 Julia Velikina 2.2 Design of Compressed Sensing Reconstruction for Highly Accelerated Time-Resolved MR Angiography 11:24 Yijing Wu 2.3 Low Dose HYPR FLOW 11:36 Lan Ge 2.4 Myocardial Perfusion MRI in Canines with Improved Spatial Coverage, Resolution and SNR 11:48 Nicole Seiberlich 2.5 Reconstruction of MR Angiography Images Using Gradient Descent with Sparsification 12:00 pm – 1:00 pm Lunch MR-Angioclub East Lansing 2009 11 1:00 pm – 2:45 pm Session 3 Plaque Characterization and Vessel Wall Imaging Session Chairs: J. Kevin DeMarco, Brian K. Rutt 1:00 Chun Yuan 3.1 Carotid Plaque Imaging and Clinical Risk Assessment 1:12 Hideki Ota 3.2 Carotid Intraplaque Hemorrhage Is Associated with Enlargement of Lipid-rich Necrotic Core and Plaque Volume Over Time: In Vivo 3T MRI Prospective Study 1:24 David Zhu 3.3 The 3D SHINE Sequence Optimizes the Quantification of Carotid Intraplaque Hemorrhage 1:36 Jinnan Wang 3.4 Improve Intraplaque Hemmorhage Detections by a Phase Sensitive IRTFE (SPI) Sequence 1:48 William Kerwin 3.5 Fibrous Cap Thickness Assessment: Fact or Fiction? 2:00 Rui Li 3.6 Gradient Echo Based Sequence Provides More Information from Ex Vivo Carotid Plaque Specimens 2:12 Seong-Eun Kim 3.7 Improved Black Blood Multi-Contrast Protocol for In-vivo Atherosclerotic Imaging 2:24 Yi Wang 3.8 3D peripheral vessel wall MRI with flow-insensitive blood suppression and isotropic resolution at 3 Tesla 2:36 Rock Hadley 3.9 A 16 Channel Anterior Neck RF Coil for Cervical Carotid MRA 2:45 pm – 3:15 pm 12 Coffee Break MR-Angioclub East Lansing 2009 3:15 pm – 5:00 pm Peripheral MRA Session 4 Session Chairs: Jeffrey Maki, Tim Leiner 3:15 Tim Leiner 4.1 Gadobenate dimeglumine vs. gadopentetate dimeglumine for peripheral MR angiography: comparison with DSA 3:27 Jeffrey Maki 4.2 Dose Comparison between Conventional and High Relaxivity Contrast Agents in Peripheral MRA 3:39 Matthias Voth - NOT ATTENDING 4.3 Periipheral MRA with Continuous Table (CTM) Movement in Combination with High Temporal and Spatial Resolution TWIST – MRA with 0.1 mmol/kg Gadobutrol at 3.0T 3:51 John Sheehan 4.4 Flow and Motion-Insensitive Unenhanced MR Angiography of the Peripheral Vascular System – A pilot study in the lower extremity 4:03 Clifton Haider 4.5 A Comparison of Time-Resolved 3D CE-MRA with Peripheral Run-off CTA in the Calves 4:15 Kang Wang 4.6 3D Time- Resolved MR Angiography of Lower Extremities using Cartesian Interleaved Variable Density Sampling and HYPR Reconstruction 4:27 Zhaoyang Fan 4.7 3D Noncontrast MRA Using FSD – Prepared Balanced SSFP 4:39 James Carr 4.8 Non Contrast MRA of the Hand in Patients with Raynauds disease using Flow Sensitized Dephasing Prepared SSFP 4:51 Casey Johnson 4.9 Two-Station Time-Resolved CE-MRA of the Lower Legs MR-Angioclub East Lansing 2009 13 Wednesday, September 30th, 2009 8:00 am – 9:45 am Session 5 Contrast Agents Performance and Safety Session Chairs: Thomas M. Grist, Martin R. Prince 8:00 Mark Hibberd 5.1 An Update on the Clinical Experience with Gadofosveset 8:12 Edward Parsons 5.2 A Re-analysis of MS-325 (gadofoveset trisodium) Clinical Trial Data in Support of US-FDA Approval 8:24 Manuela Aschauer 5.3 Gadofoveset Excretion into Human Breast Milk 8:36 Thomas Grist 5.4 Overview of Gd-BOPTA Phase III Trail for CEMRA: What are the results, and how do we move forward? 8:48 Guenther Schneider 5.5 Safety of Gadobenate dimeglumine (Gd-BOPTA) in Cardiovascular Imaging of Pediatric Patients 9:00 Giles Roditi 5.6 Retrospective 7 year Study of the Incidence of Nephrogenic System Fibrosis in Patients Investigated with Gadolinium Contrast-Enhanced Renal Magnetic Resonance Angiography Martin Prince 5.7 Risk Factors for NSF: a Meta-analysis 9:12 9:24 James Varani 5.8 Extracellular matrix metabolism in organ-cultured skin from patients with endstage renal disease: Response to gadolinium based MRI contrast agents 9:36 Zheng-Rong Lu - NOT ATTENDING 5.9 Manganese Based Biodegradable Macromolecular MRI Contrast Agents for Cardiovascular Imaging 9:45 am – 10:15 am 14 Coffee Break MR-Angioclub East Lansing 2009 10:15 am – 12:00 pm Session 6 Hemodynamics and Perfusion Session Chairs: Scott B. Reeder, E. Mark Haacke 10:15 Thorsten Bley 6.1 Non-invasive Trans-Stenotic Pressure Measurements with 3D Phase Contrast MRA: Validation against Endovascular Pressure Measurements in Swine 10:27 Alex Frydrychowicz 6.2 Analysis of aortic hemodynamics after treatments for coarctation using flow-sensitive 4D MRA at 3T 10:39 Christopher Francois 6.3 Flow assessment of arterial dissections using 3D radial phase contrast MR angiography 10:51 Scott Reeder 6.4 High Temporal and High Spatial Resolution Perfusion Imaging of Hepatocellular Carcinoma in the Liver 11:03 Steven Kecskemeti 6.5 Stack of Stars 4D Phase Contrast Velocimetry of the Circle of Willis 11:15 Mark Haacke 6.6 High Resolution Perfusion Weighted Imaging 11:27 Luca Marinelli 6.7 Accelerated velocity imaging using compressed sensing 11:39 Michael Markl 6.8 Wall Shear Stress in Normal and Atherosclerotic Carotid Arteries 11:51 Scott McNally 6.9 MR imaging and significance of flow reversal and carotid atherosclerosis: Initial results 12:00 pm – 1:00 pm Lunch MR-Angioclub East Lansing 2009 15 1:00 pm – 2:45 pm Head and Neck MRA Session 7 Session Chairs: John Huston, David Saloner 1:00 Winfried Willinek 7.1 4D-MRA in combination with arterial spin labeling for selective and functional information in patients with AVMs 1:12 Marco Essig 7.2 Intraindividual comparison between multislice CT and 4D TWIST MRA in the assessment of residual cerebral arteriovenous malformations – a prospective study protocol 1:24 Keiji Igase 7.3 Our Strategy for the Surgical Planning with 3T MRA in Detecting Unruptured Cerebral Aneurysms 1:36 Faiza Admiraal-Behloul 7.4 Hybrid of Opposite Contrast MR Angiography of the Brain 1:48 Ek Tsoon Tan 7.5 High Resolution Fast Inversion Recovery MRA (FIR-MRA) 2:00 Nick Zwart 7.6 3D Dual VENC PCMRA using Spiral Projection Imaging 2:12 Samuel Barnes 7.7 High Resolution Simultaneous Angiography and Venography (MRAV) with a Single Echo 2:24 Bum-soo Kim 7.8 Low Dose 3D Time-Resolved MR Angiography of the Supraaortic Artery: Correlation to High Spatial Resolution 3D Contrast-Enhanced MRA 2:36 7.9 Tae-Sub Chung Obstruction of IJV by Asymmetry of Lateral Mass of Atlas on Head and Neck CEMRA and Contrast CT 2:45 pm – 3:15 pm 16 Coffee Break MR-Angioclub East Lansing 2009 3:15 pm – 5:00 pm Thoracic MRA Session 8 Session Chairs: Winfried A. Willinek, Charles L. Dumoulin 3:15 Dipan Shah 8.1 Evaluation of Gd-DOTA (DOTAREM) enhanced MRA compared to timeof-flight MRA in the diagnosis of clinically significant non-coronary arterial disease at 1.5 and 3.0 Tesla 3:27 Mark Schiebler 8.2 Pulmonary MRA in 75 patients with dyspnea 3:39 Mark Schiebler 8.3 Origin and Frequency of artifacts in Contrast Enhanced Pulmonary MRA in 80 patients with dyspnea 3:51 Paul Stein 8.4 Gadolinium Enhanced Magnetic Resonance Angiography for Pulmonary Embolism: Results of PIOPED III 4:03 Loic Boussel 8.5 4D time-resolved MR angiography for non-invasive pulmonary postembolization AVM patency assessment 4:15 Timothy Carroll 8.6 Radial Sliding Window MRA in Pulmonary Hypertension 4:27 Peng Hu 8.7 Non-Contrast Enhanced Pulmonary Vein MRI with a Spatially Selective Slab Inversion Preparation Sequence 4:39 Grace Choi 8.8 MRA with the “No Phase Wrap” 4:51 Hitoshi Miki 8.9 Unruptured Intracranial Aneurysms; Detection and Follow-up on 3.0T MRA MR-Angioclub East Lansing 2009 17 Thursday, October 1st, 2009 8:00 am – 9:45 am Session 9 Cardiac Imaging / Coronary MRA Session Chairs: Debiao Li, Harald H. Quick 8:00 Oliver Wieben 9.1 Comprehensive PC MR Imaging in Congenital Heart Disease 8:12 Gary Liu 9.2 Ultrasound guided cardiac gating for coronary MRA 8:24 Himanshu Bhat 9.3 Contrast-Enhanced Whole-Heart Coronary MRA at 3T Using Gradient Echo Interleaved EPI (GRE-EPI) 8:36 Jingsi Xie 9.4 Feasibility of Whole-Heart Coronary MRA on 3 Tesla Using Ultrashort-TR SSFP VIPR 8:48 James Goldfarb 9.5 Cardiac Imaging: Methods for the Detection of Intramyocardial Fat 9:00 Dana Peters 9.6 3D spiral high-resolution late gadolinium enhancement 9:12 Ben Grabow 9.7 Temporal Filtering for Sliding Window Time-resolved Angiography; Beyond Density Compensation Solutions 9:45 am – 10:15 am 18 Coffee Break MR-Angioclub East Lansing 2009 10:15 am – 12:00 pm Abdominal MRA Session 10 Session Chairs: Franz Ebner, Walter F. Block 10:15 Kevin Johnson 10.1 Angiographic and Hemodynamic Assessment of the Hepatic Vasculature in Portal Venous Hypertension using High Resolution PC VIPR 10:27 Guenther Schneider 10.2 Renal MR angiography: multicenter intraindividual comparison of gadobenate dimeglumine and gadofosveset trisodium 10:39 Manojkumar Saranathan 10.3 FINESS (Flow Inversion-prepared Non-contrast Enhancement in the Steady State): A novel technique for non-contrast renal MRA 10:51 Tiffany Newman 10.4 Magnetic Resonance Angiography of the skin for perforator-based autologous breast reconstruction 11:03 Isabelle Parienty 10.5 Time-SLIP versus DSA in Patients with Renal Artery Stenosis 11:15 Katherine Wright 10.6 Simultaneous Renal Angiography and Perfusion Measurement Using Time-Resolved MRA 11:27 Nathan Artz 10.7 Assessing Kidney Perfusion using Arterial Spin Labeling and Radial Acquisition for Rapid Characterization of Inflow Dynamics 11:39 Walter Block 10.8 Imaging Capabilities for Real-time Guidance and Verification of Transcatheter Arterial Chemoembolization (TACE) Procedures 12:00 pm – 1:00 pm Lunch MR-Angioclub East Lansing 2009 19 1:00 pm – 2:45 pm Session 11 Venous Imaging and Beyond Session Chairs: Frank R. Korosec, Yi Wang 1:00 M Louis Lauzon 11.1 Non-Contrast-Enhanced MR identification of DVT 1:12 Mark Haacke 11.2 Susceptibility mapping as a means to image veins 1:24 Petrice Mostardi 11.3 Modified CAPR MRA: Improved Imaging of the Arterial and Venous Phases 1:36 Hyun Jeong 11.4 CAMERA: Contrast-enhanced Angiography with Multi-Echo and Radial kspace 1:48 Philip Robson 11.5 Time-Resolved, Vessel-Selective, Cerebral Angiography Using Arterial Spin Labelling 2:00 David Steinman 11.6 Quantifying Lumen Geometry from Routine Carotid CEMRA 2:12 Yi Wang 11.7 Magnetic Source MRI for Quantitative Brain Iron Mapping 2:24 Kheireddine El-Boubbou 11.8 Targeted Glyco-Magnetic Fe304 Nanoprobes for Detections and Molecular Imaging of Atherosclerosis 2:45 pm – 3:15 pm 20 Coffee Break MR-Angioclub East Lansing 2009 3:15 pm – 5:00 pm Session 12 New Horizons and Challenges in MRA Session Chairs: E. James Potchen, Charles A. Mistretta 3:15 James E. Siebert Presentation of the Best Student Posters Awards 3:20 Charles Mistretta 12.1 4D DSA and Fluoroscopy: A New Challenge for MRA? 3:32 Bas Versluis 12.2 MR Angiography of muscular and collateral arteries in peripheral arterial disease: reproducibility of morphological and functional vascular status 3:44 Matt Bernstein 12.3 Multicenter Studies: Lessons Learned from ADNI 3:56 David Saloner 12.4 Imaging Considerations in Serial Studies of Vascular Disease 4:08 Mark Ladd 12.5 Towards Abdominal MRA at 7 Tesla 4:20 Harald Quick - NOT ATTENDING: Mark Ladd to present 12.6 7 Tesla Cardiac MRA in Humans 4:32 George Abela 12.7 The Role of Cholesterol Crystals in Acute Cardiovascular Events: Identifying the Cause for Gender Differences in Clinical Presentation 4:44 James E. Potchen Perspective on MRA and the MR Angio club (no abstract) 4:56 Kevin DeMarco Presentation of the new President of the MRA Club Announcement of the new President Elect Announcement of the 22nd International MRA Conference 2010 in Seoul, South Korea MR-Angioclub East Lansing 2009 21 Posters MRA Methods P1 Manuela Aschauer CE-MRA with tailored 3D random sampling patterns and nonlinear parallel imaging reconstruction P2 Jason Mendes Handling Motion in Sparse MRA with Whiskers P3 Jordan Hulet Improved Carotid Imaging with HASTE using a reduced FOV and increased gradient performance P4 Randall Stafford Towards Continuously Moving Table NCE Peripheral MRA P5 Matthew Latourette R2* Calibration Phantoms for Cardiovascular Studies P6 Giles Roditi Pictorial Review of Supra-Aortic Artery Pathologies as Visualised with MRA using Blood Pool Contrast Agent P7 Kristine Blackham Robust Clinical Application of Time-Resolved MRA P8 Jonathan Suever Reproducibility of Aortic Pulse Wave Velocity Measurements Obtained with Phase Contrast Magnetic Resonance (PCMR) and Applanation Tonometry P9 Gregory Wilson Motion-compensated, flow-independent, non-contrast-enhanced renal MR angiography MRA Applications 22 P10 Chang-Ki Kang Vascular response during visual stimulation at 3T MRI: functional phase contrast angiography (fPCA) study P11 Jongmin Lee Optimization of Phase-contrast MR-based Flow Velocimetry and Shear Stress Measurement P12 John Oshinski Blood Flow Patterns in the Abdominal Aorta of Mice: Implications for AAA localization MR-Angioclub East Lansing 2009 Posters Plaque Characterization and Vessel Wall Imaging P13 Niranjan Balu 3D Vessel Wall imaging of multiple vascular beds P14 Keigo Kawaji Feasibility Study of Combining 3D SSFP with T2prep inversion Recovery (T2IR) for Black Blood Vessel Wall Imaging P15 Zhaoyang Fan Identification of Optimal First-Order Gradient Moment for Flow-Sensitive Dephasing (FSD) Preparation P16 Rahul Sarkar Combined Segmentation of Lumen and Intraplaque Hemorrhage in Black-blood T1-Weighted Carotid Imaging P17 Rahul Sarkar Automatic Registration of Multiparametic T1 Weighted Images Using FOVSelective Mutual Information CFD Modeling and MR-Guided Endovascular Interventions P18 Haruo Isoda MR fluid dynamics using 4D-Flow for intracranial aneurysms with growing blebs and a ruptured intracranial aneurysm P19 Charles Dumoulin Phase-Field Dithering for Active Catheter Tracking P20 Ethan Brodsky Interventional Device Tracking and Imaging Using an Extensible Real-Time System P21 Mahdi Salmani Rahimi Simplified Catheter-based Multimode Coil for Active MR Tracking and Intravascular Imaging P22 Krishna Kurpad Transmit Power Optimization for Tracking, Wireless Marker and Imaging applications of a Multi-mode Endovascular coil MR-Angioclub East Lansing 2009 23 1.1 Non-CE imaging of the pulmonary arteries Liesbeth Geerts, Marco van Essen, Gregory Wilson, Tomoyuki Okuaki Philips Healthcare, Best, The Netherlands Purpose Bright blood ASL in a single acquisition – without subtraction of tag-on and tagoff acquistions – can be used to depict vascular structures [1].The contrast mechanism relies on inflow of fresh spins into the imaging region. Therefore, optimal imaging parameters, such as the time allowed for inflow, may be patient dependant. The purpose of this study was to 1) evaluate the use of a bright blood ASL approach in a single acquisition for depiction of the pulmonary vasculature and to 2) evaluate the effect of different TI times. Methods A combination of a spatially non-selective and a selective inversion pulse was implemented on a 1.5T Achieva scanner to obtain a bright blood ASL in a single acquisition. A respiratory triggered 3D-TSE was used for readout. 12 Healthy volunteers (mean 54.2 years, range 37 to 71) were scanned, using TI’s of 500, 800 and 1100 ms. In two volunteers, additional measurements at TI 300, 700 and 900 ms were performed. Image quality was qualitatively rated. Results A TI of 800 ms gave the best depiction of a pure arterial filling. At longer TI (1100 ms) filling of the distal segments has progressed, however, also the venous signal becomes more apparent. TI 500 ms TI 800 ms TI 1100 ms Conclusion Bright blood ASL in a single acquisition is suitable to depict the pulmonary vasculature. A TI of 800 ms was found to give good depiction of the pulmonary arteries. [1] Miyazaki. Radiology 248(1);20-43 (2008) 24 MR-Angioclub East Lansing 2009 1.2 Accelerated Time Resolved Inflow with 3D Radial bSSFP Kevin M. Johnson, Oliver Wieben, Patrick Turski, Charles Mistretta Departments of Medical Physics and Radiology, University of Wisconsin, Madison, WI, USA Purpose: Recently the combination of blood tagging schemes [1] and bSSFP acquisitions has allowed for sub second resolution of vascular filling dynamics [2]. However, bSSFP can be sensitive to artifacts from off-resonance banding and flow artifacts. In this work we investigate the use of 3D radial trajectories for highly accelerated non-contrast enhanced angiography with reduced flow and banding related artifacts. Methods: All imaging was performed on a clinical 1.5T scanner. An ecg-triggered, inversion-recovery prepared, cardiac interleaved, bSSFP, 3D sequence was implemented with both dual-half echo (non-flow compensated) and 4-half radial trajectories (flow compensated) [3]. Angiographic images are acquired by subtracting of a pass with nonselective inversion pulse from a pass with a selective inversion just covering the imaging volume. Typical imaging parameters for 4-half echo (2-half echo) are TR=4.0ms (3.0ms), readout bandwidth=+/-125kHz, FOV=24x24x14cm, resolution= 0.94x0.94x0.94mm, flip angle = 45º, RR interval = 2, total scan time =~6min for 45,000 (60,000) total unique projections. Results: Representative images from the 4-half echo sequence are shown in Figure 1. Images show filling of the major vessels without substantial blurring. Signal diminishes as the signal recovers from inversion and as blood distant from the volume fills the vessels. Results indicate improved performance of the 4-half echo trajectory, due to flow compensation and improved sampling efficiency. Conclusion: The proposed 3D radial sequence provides significantly shorter TRs for a given resolution than their Cartesian counterparts. Considerable acceleration can be achieved allowing for higher-resolution imaging or reduced scan time. References: 1. Kim SG. MRM 34:293-301 (‘95) 2. Bi et al. Proc MRA Club 08 pg.31 3. Lu et al. MRM 53:692-699 (05’) MR-Angioclub East Lansing 2009 25 1.3 Dose 4D MR Angiography Gerhard Laub, Ph.D. Siemens Healthcare, USA Time resolved contrast-enhanced MR angiography has been increasingly used to evaluate the hemodynamic status of normal versus abnormal vasculatures. Fast imaging sequences, parallel imaging, and view-sharing techniques have been applied to provide the needed temporal and spatial resolution. The Gadolinium-based contrast agent is sometimes injected in double dose to enhance the image quality. In light of NSF and the desire to lower the amount of Gadolinium-based contrast agent to the patient, we have investigated the use of time resolved TWIST imaging (Time-resolved Imaging with Stochastic Trajectories) in combination with a small dose of diluted contrast agent for 4D imaging of the extracranial vasculature. In this study, we tested the feasibility of using a very low dose of contrast for timeresolved MRA in patients referred to get a clinical MRA examination. For low dose dynamic MRA, 1-2 ml of Gd-DTPA, diluted with saline at a rate of 1 part Gd and 3 parts of saline, was injected at a rate of 2 ml/sec. This was compared to routine contrast-enhanced MRA using a single dose (0.1 mmol/kg) of contrast agent. All imaging was performed at 3T using a combination of a 12-channel head array, 4-channel neck array, and 6-channel thorax array to extend the FOV and cover the entire aortic arch, the carotid arteries, and the intracranial arteries all dynamically. Parallel imaging was used in two phase encode directions with an acceleration factor of up to 9. An additional acceleration factor of 3.8 was achieved using the TWIST dynamic mode. 3D imaging, with 100 slices (slice resolution = 2.5 mm), was acquired with an in-plane resolution of 1.3 mm x 2.2 mm and interpolated to isotropic voxels of 1.3 mm. Using only 2 ml of contrast agent, or less, all time-resolved results were clinically useful to provide functional information in addition to the anatomical information provided by the high-resolution, single-phase contrast-enhanced MRA. In this preliminary study, there was good agreement between the low dose, time-resolved MRA and the routine, high-resolution contrast-enhanced MRA. By using a combination of parallel imaging and the TWIST dynamic mode, the temporal update rate was under 2 sec for each 3D volumetric data set, depending on spatial resolution and vessel coverage. Time-resolved, three-dimensional MRA with near isotropic resolution and large coverage is feasible using a small amount of a Gadolinium-based contrast agents. Further studies involving larger number of patients are needed to determine whether very low dose time-resolved MRA would lead to any difference in clinical diagnosis. 26 MR-Angioclub East Lansing 2009 1.4 Balanced-gradient TSE for Non-contrast Peripheral MRA 1 1,2 2,3 2,3 1,2 SW Fielden , JP Mugler III , KD Hagspiel , CM Kramer , and CH Meyer 2 Biomedical Engineering, University of Virginia, Charlottesville, Virginia, United States, Department of Radiology, University of Virginia, Charlottesville, Virginia, United States, 3Department of Medicine, University of Virginia, Charlottesville, Virginia, United States 1 Purpose: To develop a 3D balanced-gradient TSE sequence for generating non-contrast, non-subtractive, peripheral angiograms. Methods: A 3D dataset was acquired on each of 9 healthy volunteers on a Siemens 3T Trio scanner (Siemens Medical Solutions) with a TSE sequence modified with (1) a frequency-encoding gradient which is rewound between each set of refocusing pulses and (2) excitation and refocusing RF pulses applied along the same axis with a 180° phase alternation along the echo train, analogous to common implementations of balanced SSFP. Sequence parameters were as follows: TR/TE/Echo Spacing = 3000/230/3.1 ms, resolution = 1.4 x 1.4 x 1.5 mm3, acquisition time = 5.38 ± 0.50 min. Image quality was assessed based on SNR, CNR, and vessel sharpness. Results and Conclusion: The incorporation of balanced gradients and RF-pulse phase alternation in a TSE sequence resulted in T2-weighted images with reduced flow-related signal loss and blurring, yielding good contrast between arteries and veins without subtraction1. As seen in the coronal MIP figure to the right, blood-muscle contrast is similar to that obtained with flow-independent angiography2 and to that obtained with traditional TSE sequences. The residual fat and edematous signal may be mitigated in the future via chemically selective pulses and an inversion preparation. References: 1 Miyazaki M, et al. Radiology, 227(3):890-6, 2Brittain J, et al. Magn Reson Med, 38(3):343-454 MR-Angioclub East Lansing 2009 27 1.5 Self-Gated Free Breathing 3D Cardiac Cine Imaging With Data Acquisition During Slice Encoding 1 1 1 Jing Liu , Martin R. Prince , Yi Wang Weill Medical College of Cornell University, New York, NY Email: [email protected] 1 Self-gated free-breathing 3D cardiac cine imaging is becoming a promising technique for heart visualization and functional measurements. The common techniques for self-gating requires additional scan time for acquiring extra gating data. The purpose of this study was to develop a cardiac and respiratory self-gated free-breathing 3D pulse sequence that avoids extra scan time and provides robust self-gating information. The multi-echo 3D hybrid radial pulse sequence was modified by adding readout during the slice encoding pulse. The acquired data during slice encoding pulse are phase encoded k-space centers along kz (Fig. 1). Different slice encoding gives different kz coverage, while the combination of sequential slice encodes gives full kz coverage. The 1D Fourier transform of this combined data provides the z intensity profile, which Fig. 1. Samples acquired during sequential slice encodes (10 slices for example). contains both cardiac and respiratory motion information for cardiac and respiratory self gating. Typical imaging parameters were: TR/TE/FA/BW/FOV=4.5ms/1.3ms/40°/±125kHz/34cm, 256x256 matrix, 10 slices of slice thickness 10mm. A 1.5 T GE EXCITE 14M4 scanner was used. Fig. 3 shows short axis heart images acquired within a scan time of 3.4 minutes. The respiratory gating efficient was 50%. A self-gating technique with during extra slice data acquisition encoding was demonstrated with 3D hybrid radial imaging. It can be applied to other 3D pulse sequences, such as 3D Cartesian and spiral pulse sequences. Further validation experiments of the Fig. 3. Short axis images of five presentative slices at end diastole (top row) and end systole (bottom row). proposed self-gating method are underway. 28 MR-Angioclub East Lansing 2009 2.1 A Simple View of Compressed Sensing and How it Could Change Everything We Do in MRI and MRA. 1 1 Mark A Griswold , Nicole Seiberlich Dept. of Radiology, Case Western Reserve University, Cleveland, OH 1 MRA provides exquisite depiction of vascular abnormalities without the ionizing radiation found in conventional DSA or CTA. However, significant drawbacks still exist. This is primarily due to the limited speed and SNR of MRI, and most important, the fact that they are linked to each other. Traditionally, any increase in imaging speed has required a loss in SNR and vice versa. In order to realize any truly dramatic increases in either SNR or imaging time, some way to break this relationship must be found. The class of newly described Compressed Sensing (CS) methods promises to revolutionize MRI by breaking this link and could potentially allow the development of a set of completely new imaging strategies with dramatic increases in SNR and imaging speed. Unfortunately much of the field of CS is described in highly mathematical terms, limiting the ability of the average listener to fully grasp the concepts involved. In this talk, we will review some of these basic concepts of CS in a non-mathematical, intuitive way. In general, we will demonstrate how CS will change the focus of an MR acquisition from simply collecting images to directly collecting information. In particular, we will focus on how somewhat older methods, such as UNFOLD, BLAST, PARSE, and other related methods can be seen in the framework of CS, and will also show how methods such as HYPR meet this goal of directly collecting the important information. Finally, we will highlight several examples where these new methods have dramatically improved both the SNR and temporal resolution beyond all previously established limits to provide clinically useful exams in dramatically reduced time with increased SNR. MR-Angioclub East Lansing 2009 29 2.2 Design of Compressed Sensing Reconstruction for Highly Accelerated Time-Resolved MR Angiography Julia Velikina, Alexey Samsonov Departments of Medical Physics and Radiology, University of Wisconsin – Madison Introduction: Many angiographic tasks require high temporal resolution, which results in data undersampling. A number of constrained reconstruction methods including compressed sensing (CS) have been proposed to mitigate problems of conventional reconstruction (aliasing and low SNR)1,2. However, even natural sparsity of angiographic images cannot support acceleration factors higher than 4-8 without loss of spatial resolution and artifacts. In time-resolved imaging, the efficiency of CS can be extended by promoting joint sparsity in both spatial and temporal dimensions3,4. In this work, we evaluate the performance of two CS approaches: k-t FOCUSS5 (sparsity in x-f domain) and temporal constraining (sparsity of finite differences in x-t domain) in contrastenhanced (CE) and phase contrast (PC) imaging. Theory and Methods: We adapted spatial/temporal regularization4 into CS framework. The time series f is estimated as follows: 2 f = arg min Ef − m + λΦ( Lf ) 2 f vector m , regularization parameter Φto be a hybrid 1 acceleration (via / 2 norm 6 [1], with the encoding matrix E . L is a 1 λ st , data or 2nd temporal derivative. We chose to provide both SNR optimization (via 2) and CS 1norm). Results and Conclusions: All approaches were tested on CE and PC data acquired in healthy volunteers. For Cartesian trajectories acceleration of 8-10 were achieved, while radials allowed acceleration up to 40 without loss of temporal and spatial resolution. The developed temporal regularization was found to outperform kt FOCUSS in preservation of temporal waveforms for high acceleration factors. Additionally, it was found more forgiving to patient motion than kt FOCUSS. CS methods based on sparsity in x-t domain (Eq. [1]) are promising to accelerate time-resolved angiography. References: [1] Çukur T, et al MRM, 2009;61:1122. [2] Mistretta CA, et al. MRM 2006;55:30. [3] Portniaguine, et al. ISMRM 2003,481. [4] Samsonov, ISMRM 2005, 2311. [5] Tsao J, et al. MRM 2003;50:1031. [6] Bube KP, et al. Geophysics,1997:62:1183. 30 MR-Angioclub East Lansing 2009 2.3 Low Dose HYPR FLOW Yijing Wu, Steven Kecskemeti, Kevin Johnson, Charles Mistretta, Patrick Turski Departments of Medical Physics and Radiology, University of Wisconsin, Madison, WI INTRODUCTION: Time resolved contrast-enhanced magnetic resonance angiography has been widely used to evaluate the hemodynamics of the vascular structure. Due to the recent concern of NSF disease, eliminating or reducing the Gadolinium-based contrast agent is more desirable than ever. Phase Contrast (PC) HYPR FLOW is able to decouple the high spatial resolution and SNR, which require relative long scan time, from high temporal resolution, which demands for fast data acquisitions, and used HYPR constrained reconstruction to obtain a time series of images with both high temporal resolution, isotropic high spatial resolution, high SNR and quantitative flow dynamics from the PC images. Our hypothesis is that the SNR of the HYPR FLOW images depends on the post contrast PC composite, which requires minimum amount of contrast agent. High temporal and spatial resolution time resolved contrast-enhanced MRA can be obtained by using low dose HYPR FLOW method with reduced contrast agent. METHODS AND RESULTS: Low dose HYPR FLOW has been tested in three normal volunteers and one brain AVM patient. Contrast agent was reduced to one half or one quarter of a single dose (0.1 mmol/kg). Following contrast injection, a CE-MRA examination of the head is performed using time resolved milt-echo VIPR. Subsequently, a 5 minute PC VIPR acquisition was acquired and then used as a composite image for HYPR LR reconstruction. The figure on the right shows a time series of HYPR FLOW 1 2 4 6 25 35 images of the AVM patient with half dose contrast agent (~ 6 cc). Numbers on 8 15 the images are relative time frames after contrast arrival. Frame rate was 2/s, with 0.7 mm isotropic spatial resolution. Our preliminary results show that low dose HYPR FLOW is able to provide high temporal and spatial resolution with adequate SNR and potential shorter bolus dispersion for intracranial MRA. MR-Angioclub East Lansing 2009 31 2.4 Myocardial Perfusion MRI in Canines with Improved Spatial Coverage, Resolution and SNR 1 Lan Ge , Aya Kino1, Daniel Lee1, Rohan Dharmakumar1, Mark Griswold2, Charles Mistretta3, James Carr1, Debiao Li1 1 Northwestern University, Chicago, IL, USA, 2 Case Western Reserve University, Ohio, USA, 3 University of Wisconsin-Madison, Madison, WI, USA Purpose: First-pass perfusion MRI is a promising technique for detecting ischemic heart disease. A combination of sliding window and CG-HYPR (1, 2) methods (SW-CG-HYPR) have been proposed to increase spatial coverage, resolution, and SNR(3). In this work, using a controlled animal model, we compare this new method with conventional clinical protocols and test whether the flow deficits can be detected accurately. Methods: Five dogs with LCX occlusion were scanned using a 1.5T system during the first-pass of the contrast agent in stress condition. An ECG-triggered, turbo-FLASH sequence with radial k-space sampling and saturation recovery (SR) preparation was used in this study. CG-HYPR method was used to reconstruct the time-resolved images. The signals from the left ventricle, healthy myocardium, and flow deficits for all of the three methods were measured and compared. Results: The SNR of the left ventricle at peak enhancement with SW-CG-HYPR (32.1±2.32) is significantly higher than turbo-FLASH (20.6±2.62) and EPI (12.9±1.95). Figure 1 shows examples of the comparison. The defects caused by LCX occlusion can be clearly delineated in SW-CG-HYPR images. The signal intensity changes of the healthy myocardium are highly correlated with a correlation coefficient of 0.956. Conclusions: In conclusion, SW-CG-HYPR is a turbo-FLASH vs. SW-CG-HYPR EPI vs. SW-CG-HYPR Reference images promising method to improve the Deficits SW-CG- Deficits myocardial perfusion MRHYPR imaging with reduced images acquisition window Figure 1. Comparison of conventional method and SW-CG-HYPR increased spatial coverage, improved spatial resolution and SNR. References: 1. Mistretta CA, et. al. MRM, 55: 30-40, 2006. 2. Griswold MA, et. al. Proc ISMRM, Berlin, 2007: 834. 3. Ge L, et al. Proc ISMRM, Toronto, 2008: 43 32 MR-Angioclub East Lansing 2009 2.5 Reconstruction of MR Angiography Images using Gradient Descent with Sparsification Nicole Seiberlich and Mark A. Griswold Department of Radiology, Case Western Reserve University, Cleveland OH Purpose: Recently, a new method for generating sparse images from highly undersampled data, Gradient Descent with Sparsification, has been introduced [1]. This method iteratively determines x, the sparse image to be reconstructed using the following formulation: x ← H ( x + γ ⋅ ΦT ( y − Φx)) where y is the k-space data,Φ is the resampling/degridding operation, H is a thresholding operation, and γcontrols the rate of convergence. This simple reconstruction algorithm was tested for the reconstruction of images from highly undersampled MR Angiography data. Methods: CE-MRA data were acquired on a patient with an arterial-venous malformation (AVM) with the following parameters: radial GRE acquisition, TR=3ms, TE=1.5ms, total projections=1344, 75% asymmetric echo, Matrix=192x192, FOV=220x220 FA=20°, Partitions=15, slice thickness=4mm. Images were reconstructed using Gradient Descent with Sparsification with 32 projections per frame, with the thresholding value H determined using the center of k-space and γ fixed as suggested in [1]. Results: Time frames showing the arrival of contrast into the AVM reconstructed Descent with using Gradient Sparsification are shown in Figure 1. The streaking Figure 1: Gradient Descent with Sparsification reconstruction depicting flow of contrast into the AVM. commonly evident in such highly undersampled images has been removed, and the vessels feeding the AVM are clearly depicted without venous contamination. Conclusion: Gradient Descent with Sparsification is a simple and fast method to reconstruct highly undersampled sparse images, including MRA data. Unlike other reconstruction methods, no composite image is required, reducing reconstruction errors. Further improvements to the method such as the inclusion of coil sensitivity maps or constraint on the thresholding images will be considered in the future. References: [1] Garg R and Khandekar R. Proc. 26th International Conference on Machine Learning, Montreal, Canada, 2009. MR-Angioclub East Lansing 2009 33 3.1 Carotid Plaque Imaging and Clinical Risk Assessment Chun Yuan, PhD, Hunter Underhill, MD, Thomas Hatsukami, MD Vascular Imaging Laboratory, School of Medicine, University of Washington, Seattle, WA Carotid MRI has been proven to be able to measure plaque size and characterize tissue composition. This information provides unique opportunities to study the ‘vulnerable plaque’ in vivo. This talk aims to summarize the current understanding of vulnerable plaque features based on longitudinal studies of carotid atherosclerosis, as well as the relationship between carotid atherosclerosis and neurological symptoms. This talk will also discuss the current technical and clinical needs of atherosclerosis imaging. 34 MR-Angioclub East Lansing 2009 3.2 Carotid Intraplaque Hemorrhage Is Associated with Enlargement of Lipid-rich Necrotic Core and Plaque Volume Over Time: In Vivo 3T MRI Prospective Study H Ota1, D. Zhu1M, JK DeMarco1 Michigan State University, East Lansing, MI, USA 1 Purpose: To test the hypothesis that intraplaque hemorrhage identified by 3D Inversion recovery fast SPGR (IR-FSPGR) images contributes to plaque progression over time as measured by multi-contrast carotid MR imaging at 3.0T. Methods: Twenty-six consecutive subjects with known 50-99% carotid stenosis underwent serial carotid 3T MRI scans with a multicontrast weighted protocol (pre- and post-contrast T1W, T2W, 3D TOF and 3D IRFSPGR) with intervals of 0.5-2.2 years. Two reviewers blinded to subject’s clinical information and scan date interpreted images. In order to insure a similar coverage of the carotid artery for quantitative measurements, only image locations that could be matched across 2 time points were reviewed. The volumes of vessel wall, lumen, lipid-rich necrotic core, hemorrhage and calcification were documented for each study. The changes in each metric adjusted for the period (mean ± standard deviation/year) were evaluated in the groups with and without intraplaque hemorrhage at the baseline scans using one sample t-test. Results: Eight subjects (31%) had intraplaque hemorrhage at baseline. Mean lipid-rich necrotic core volume at baseline was significantly larger for patients with hemorrhage (p=0.001) than those without hemorrhage. The mean volumes of lumen, wall and calcification at baseline were not significantly different. In the eight subjects with intraplaque hemorrhage at baseline, significant increases of the volumes were found for wall (23.7± 26.3mm3/year, p=0.038), lipid-rich necrotic core (44.0±30.6 mm3/year, p=0.005), and intraplaque hemorrhage (24.1±23.4 mm3/year, p= 0.022). The changes in the volumes of lumen and calcification were not significant. In the remaining 18 subjects without intraplaque hemorrhage at baseline, the changes in the volumes of wall, lipid-rich necrotic core were not significant over time. Conclusion: Intraplaque hemorrhage noted on baseline carotid wall 3T MR imaging is associated with larger lipid rich necrotic core at baseline as well as significant progression of overall plaque volume, the size of the lipid-rich necrotic core and intraplaque hemorrhage over time. MR-Angioclub East Lansing 2009 35 3.3 The 3D SHINE Sequence Optimizes the Quantification of Carotid Intraplaque Hemorrhage 1 David C. Zhu , Hideki Ota1, Marina S. Ferguson2, Anthony T. Vu3, J. Kevin DeMarco1 1 Michigan State University, 2University of Washington, 3GE Healthcare Introduction: Carotid intraplaque hemorrhage (IH) has been shown to promote plaque progression. A novel optimized 3D SHINE sequence was developed to detect IH based on high T1 contrast and to characterize hemorrhage into type I (early) and type II (recent) based on T2* maps (1). With these advantages, algorithms were developed to estimate the size and composition of intraplaque hemorrhage. Methods: 3D SHINE images (1) were collected from nine patients with carotid IH at 3T. At the first TE, the signal intensity ratios of IH versus lumen, wall and surrounding muscle were 7.94 ± 3.4, 3.20 ± 1.21 and 3.82 ± 0.92. These high contrasts allowed for reliable computerized segmentation. A target ROI contained IH and a background ROI contained the surrounding muscle. At each slice, each voxel was normalized by the mean signal intensity of the background ROI at this slice. Normalization reduced the signal intensity variation due to the surface coil, and allowed a uniform cutoff of 1.5 on each slice to identify the probable IH voxels. These voxels were further labeled as “IH” if they were within a minimal “IH” size of 0.68 mm3 (14 voxels). The IH region was further divided into type I and type II based a T2* cutoff value of 14 ms (1). The sizes of the segmented intraplaque hemorrhage and its sub-regions were then calculated. Results and Discussion: The segmented intraplaque hemorrhage (middle figure, red/yellow: type I, blue: type II) shows a close match with the original data (left figure) and a reasonable correspondence with matched histology data (right figure). A semiautomated quantification of intraplaque hemorrhage would allow the monitoring of plaque progression and could be used as an objective tool in multi-site clinical trials. Reference:1. Zhu DC, Ota H, Vu AT, DeMarco JK. ISMRM 2009, Honolulu, Hawaii. 36 MR-Angioclub East Lansing 2009 3.4 Improved Intraplaque Hemorrhage Detection by a Phase Sensitive IRTFE (SPI) Sequence Jinnan Wang 1, Marina Ferguson 2, Chun Yuan 2, Peter Börnert 3 Affiliations: 1. Philips Research North America, 2. University of Washington, 3. Philips Research Europe Purpose: Hemorrhage in atherosclerosis plaque has been linked to an increased risk of plaque progression and rupture [1]. However, only a small portion of the dynamic range have been utilized [2, 3] by current T1 based hemorrhage detection sequences. This can lead to inaccurate identification and segmentation of plaque components. In this study, a Slab-selective Phase-sensitive IRTFE (SPI) sequence is proposed to improve image contrast of intraplaque hemorrhage. Methods: The SPI sequence is adapted from a regular IR-TFE sequence by making the inversion pulse slab selective and phase sensitive. Simulations are used to evaluate the signal contrast increase by using this novel SPI sequence. 9 carotid endarterectomy (CEA) specimens were scanned with both SPI and regular IRTFE sequences. The hemorrhage to lumen contrast for each specimen was calculated and compared between SPI and IRTFE images. Three of the nine CEA specimens were processed for histology with 1 specimen confirmed to contain intraplaque hemorrhage. Results: Simulations indicate that contrast between hemorrhage and normal vessel wall can be increased by over 50%. The hyper intense regions on both images Fig.1 Ex vivo MR images using routine IRTFE (left) and SPI (middle). Matching histology (right) confirms the presence of intraplaque hemorrhage. Note the improved contrast of hemorrhage on the SPI correspond very well sequence (arrows). to the region of hemorrhage on matching histology. An approximate 30% improved tissue contrast was obtained on the SPI image. Conclusion: Simulation and ex-vivo experiments confirm that the novel SPI technique can improve hemorrhage contrast on MR images by utilizing phase information. This improved tissue contrast can potentially improve the identification and quantification of intraplaque hemorrhage by carotid plaque MRI. References: 1. Takaya N, et al. Stroke. 2006; 37:818-23. 2. Moody AR, et al. Circulation 2003;107;3047-3052. 3. Zhu DC, et al. 2008;26:1360-6. MR-Angioclub East Lansing 2009 37 3.5 Fibrous Cap Thickness Assessment: Fact or Fiction? William S. Kerwin, Huijun Chen University of Washington, Department of Radiology, Seattle, WA, USA Purpose – Determining fibrous cap thickness in carotid atherosclerotic disease is the holy grail of plaque imaging as cap thickness may be the most important factor in distinguishing between vulnerable and stable lesions. While histological studies have indicated that the key break point for vulnerable cap thickness is on the order of 200 microns [1], MRI resolution in carotid studies is typically >500 microns. Nevertheless, claims of identifying thin caps (<250 microns) [2] and actual measurement of cap thickness [3] have been made. The purpose of this study is to assess the extent to which cap thickness can be assessed at the fringes of MRI resolution. Methods – A series of physical and computational phantoms that mimicked fibrous cap geometry were used to probe the relationship between cap thickness, image resolution, and appearance under MR imaging conditions. Based on these phantoms, a linear Fisher classifier was constructed that produced a mathematical model for classifying fibrous cap thickness derived from 11 intensity-related parameters. This classifier was then applied to 26 contrast-enhanced plaque images from 7 individuals with histologically measured cap thickness. Results – The experiments indicated that the fibrous cap could not be visualized when its thickness was less than about half the fundamental resolution, which coincides well with observations using the absence of a band separating the core from the lumen to define “thin” caps [2]. Above this threshold, differences in cap appearance that correlated with cap thickness could be ascertained even below the fundamental resolution; however, measured thicknesses were heavily biased and sensitive to SNR and physical MR parameters when the thickness was <1.5 times the fundamental resolution. The determinant from the linear Fisher classifier was significantly correlated (p<0.05) with measured cap thickness for in vivo imaging. Conclusion – MRI is able to differentiate caps of different thicknesses even below the fundamental resolution, using intensity criteria. Measurements of cap thickness will be unreliable, however, unless resolutions of 100-300 microns can be attained in vivo. [1] Redgrave Stroke 2008; [2] Hatsukami, Circ 2000; [3] Sadat, Atheroscl 2009 38 MR-Angioclub East Lansing 2009 3.6 Gradient Echo Based Sequence Provides More Information from Ex Vivo Carotid Plaque Specimens 1 1 Rui Li , Chun Yuan Affiliation: 1. Radiology Department, University of Washington Purpose: Spin echo (SE) based MRI sequences are generally used for high resolution plaque imaging. Criterion for vulnerable plaque also uses SE based T1, T2 and PD weighted images to access the components in the plaque[1]. However gradient echo (GE) based sequences has rarely been explored for plaque evaluation. This study aimed to find out whether we can get more information from GE by carotid specimen examination. Methods: Conventional GE imaging was performed on a 3T MRI system (Philips Achieva) with the carotid specimens fixed in 70% formalin solution. The main parameters were TR: 721ms, TE: 9.2ms, FA: 60o, SLC: 32, TH: 1mm, FOV: 24*24mm2, and Res: 0.16*0.16mm2. Four carotid endarterectomy specimens were inspected. Histogram, joint entropy and conditional entropy[2] were used to evaluate the information for the whole plaque and the lipid rich necrotic core (LRNC). Results: Ex vivo images and their Figure 1. Cross-sectional image and histogram of different ROI histograms are shown in figure 1. The intensity distribution of GE and PD is scattered. Information measured by joint entropy and conditional entropy is shown in table 1. PD and GE images contributed more information to joint Information H(T1, T2, PD, GE) H(T1 | T2, PD, GE) H(T2 | T1, PD, GE) H(PD | T1, T2, GE) H(GE | T1, T2, PD) Plaque 13.37 0.65 0.80 1.10 0.92 LRNC 11.31 0.29 0.30 0.49 0.47 Table 1. Joint and conditional entropy entropy. We think this extra information comes from T2* weighting in GE based sequence. Conclusion: Gradient echo based sequences can provide additional information to existing plaque imaging protocols. Reference: [1] C. Yuan, W.S. Kerwin, JMRI, 2004, 19, 710-719. [2] R.G. Gallager, Information Theory and Reliable Communication, Wiley, 1968. MR-Angioclub East Lansing 2009 39 3.7 Improved Black Blood Multi-Contrast Protocol for In-vivo Atherosclerotic Imaging 1,3 Seong-Eun Kim , John Roberts, Gerald S. Treiman2,4, Dennis L. Parker1,3 Utah Center for Advanced Imaging Research, 2VA Salt Lake City Health Care System, 3Radiology, 4 Surgery, University of Utah 1 Purpose: Atherosclerotic plaque characterization by MRI is generally based on the signal intensities and morphology of plaque in T1, PD and T2 weighted images, but intraplaque thrombus cannot be detected by conventional contrast imaging. Our multi-contrast protocol including DWI and 3D MPRAGE may help detect plaque hemorrhage and assist in the identification of plaque components in the cervical carotid artery. Methods: DW images (1.0x1.0x2.0 mm3) were acquired using 2D ss-IMIV DWEPI with b =10 and 300 s/mm2 to create the ADC maps. 2D T1w, T2w and PD images of the same locations were acquired (0.5x0.5x2.0 mm3) with 2D TSE. (The T1w with our modified version of the double inversion preparation). 3D MPRAGE images were acquired (0.5x0.5x1.0 mm3) with non-selective inversion preparation and fat saturation. Total scan time including 2D and 3D TOF was less than one hour. Results: Multi-contrast Images of atherosclerotic plaque with hemorrhage from a patient are shown at the left. The plaque area indicated by white arrows shows a moderate signal on T1w, T2w and low ADC value (0.29x10-3 mm2/s). 3D MPRAGE images show the hyperintense hemorrhage region in the carotid plaque. Discussion: The results obtained indicate that multi-contrast black blood images including DWI (as a new contrast) and 3D MPRAGE may be of substantial value for carotid plaque identification. T1w 40 T2w ADC MR-Angioclub East Lansing 2009 3DMPRAGE 3.8 3D peripheral vessel wall MRI with flow-insensitive blood suppression and isotropic resolution at 3 Tesla Thanh D. Nguyen, Keigo Kawaji, Pascal Spincemaille, Matthew D. Cham, Priscilla Winchester, Martin R. Prince, Yi Wang Weill Medical College of Cornell University, New York, NY Email: [email protected] Flow-insensitive T2-prepared inversion recovery (T2IR) has been shown to provide better blood suppression for 2D arterial vessel wall imaging at 1.5T than double inversion recovery (DIR) in the lower extremities where blood flow is slow (1). T2IR offers global blood suppression regardless of flow velocity and direction at the cost of reduced wall SNR. The purpose of this study was to develop a cardiac triggered 3D black blood T2IR fast spin echo sequence for vessel wall MRI at 3T. For comparison, 4 normal volunteers (27 ± 8 years) were imaged at both 1.5T and 3T (GE HDxt) with parameters: TR=1RR, TE=20 ms, FOV=28 cm, matrix=384x384, coronal slice thickness=1.4 mm (1.5T, interpolated to 1.5 mm) and 0.7 mm (3T), number of slices=100/200 (1.5/3T), NEX=2/1 (1.5/3T), bandwidth=±62.5 . kHz, ETL=64 (variable flip angles to increase sampling efficiency), selfcalibrated parallel imaging (ARC) factor R=1.9, partial kz factor=0.75, fat suppresion, scan time~8 min, 8channel cardiac coil. For T2IR preparation, T2PREP time=120/150 ms (1.5/3T), and TI~250/300 ms (1.5/3T). a Figure 1 shows images of distal superficial femoral and b Figure 1. Curved reformatted vessel wall images Obtained at a) 1.5T b) 3T. proximal popliteal arteries obtained at 1.5T and 3T, demonstrating extended bilateral coverage, excellent arterial blood suppression across large FOV, and good wall visualization. Note that 3T images offer true 0.7x0.7x0.7 mm3 isotropic resolution with similar wall SNR. In conclusion, the global blood suppression of T2IR and the higher SNR afforded by 3D acquisition and higher field strength allow efficient imaging of vessel wall with sub-millimeter isotropic resolution. References. Brown R. ISMRM 2009. p3844. MR-Angioclub East Lansing 2009 41 3.9 A 16 Channel Anterior Neck RF Coil for Cervical Carotid MRA Quinn Tate, Emilee Minalga, Laura C. Bell, J. Rock Hadley Utah Center for Advanced Imaging Research, Dept of Radiology, University of Utah Purpose: Increased speed in cervical carotid imaging is highly desired to reduce the artifacts from occasional motion such as swallowing. Methods: We have developed a 16 channel receive only RF coil array with overlapping elements placed on a form fitting fiberglass former as shown in Figure 1. Figure 1: A 16 channel (8 channels on each side) RF coil array for the neck, with emphasis on the cervical carotid artery. The array was fabricated based on an overlapping element philosophy. A second coil array is being fabricated based upon evaluations of this coil array. Below are shown example images from the 16 channel coil for (left) R=1 (no reduction, 6:44min), (middle) R=2 (3:25min) and (right) R=3 (2:43min) using 2D TSE with DIR prep TE/TR=8/885ms matrix=256x256 ETL=9 TI=600ms 2 AVG, FOV=130mm Results: Images shown in Figures 1 and 2 illustrate the improvement for parallel imaging over the 4 channel bilateral RF coil that we currently use. Figure 2: (left) Head/neck positioner with 4 channel (2 channel bilateral) carotid coil. (right) Comparison 2D DIR TSE images (TE/TR=8.8/885ms, TI= 600ms, 0.5x0.5mm2 inplane, 2mm slice thickness) obtained using the 4 channel coil shown at left (bottom row) and the 16 channel coil shown in Figure (top row). With a reduction factor of 2, the 16 channel coil images are less noisy. Conclusion: The new 16 channel coil shows great potential for rapid carotid MRI. 42 MR-Angioclub East Lansing 2009 4.1 Gadobenate dimeglumine vs gadopentetate dimeglumine for peripheral MR angiography: comparison with DSA SC Gerretsen1, T le Maire2, S Miller3 , SA Thurnher4, CU Herborn5, H Michaely 6, 7 8 9 10 1 H Kramer , A Vanzulli , J Vymazal , M Wasser , T Leiner 1. Maastricht University Hospital, Maastricht, The Netherlands; 2 Catharina Hospital, Eindhoven, Netherlands; 3. Eberhardt Karls University, Tuebingen, Germany; 4. Hospital Brothers of St. John of God, Vienna, Austria; 5. University Medical Center, Hamburg-Eppendorf, Germany, 6. University Hospital, Mannheim, Germany, 7. Ludwig Maximilians University, Munich, Germany, 8. Ospedale Niguarda Ca' Granda, Milan, Italy, 9. Na Homolce Hospital, Prague, Czech Republic, 10. Leiden University Medical Center, Leiden, Netherlands Purpose: To prospectively compare equivalent 0.1 mmol/kg doses of gadobenate dimeglumine and gadopentetate dimeglumine in patients undergoing contrast-enhanced MR angiography (CE-MRA) of the peripheral arteries. Methods: 96 adult subjects with suspected moderate-to-severe peripheral arterial occlusive disease (PAOD) were enrolled at 7 sites. Patients underwent 2 identical 1.5-T, 3-station, CE-MRA examinations from the aortic bifurcation to the lower leg with randomized 0.1 mmol/kg bodyweight doses of gadobenate dimeglumine and gadopentetate dimeglumine. Diagnostic performance (sensitivity, specificity, accuracy, positive predictive value [PPV], negative predictive value [NPV]) was determined in a subset of patients (n=31) that also underwent conventional DSA. The presence and extent of steno-occlusive disease on DSA images was determined on a segmental basis using a 4-point scale (1=stenosis ≤25%; 2=stenosis >25–≤50% 3=stenosis >50–99%; and 4=occlusion). Statistical analyses were performed using the Wilcoxon Signed Rank, McNemar, and Wald tests. Results: A total of 397 segments were evaluated by DSA. Of these 397 segments, 270 (68.0%) had stenoses of ≤50% while 127 (32.0%) had hemodynamically-relevant (>50%) stenoses. All 3 blinded readers reported significantly (p≤0.0017) better diagnostic performance with gadobenate dimeglumine compared to gadopentetate dimeglumine, with increases of 11.0–18.1% in sensitivity, 4.4–9.3% in specificity, and 7.8–10.1% in overall accuracy (Table 1). Readers also reported significantly (p≤0.0028) higher PPV and NPV with gadobenate dimeglumine, with differences ranging from 12.7–19.3% for PPV and 5.5–7.9% for NPV. Conclusion: In patients with suspected PAOD referred for CE-MRA, use of 0.1 mmol/kg bodyweight gadobenate dimeglumine results in significantly better diagnostic performance than use of an equivalent dose of gadopentetate dimeglumine. MR-Angioclub East Lansing 2009 43 4.2 Dose Comparison between Conventional and High Relaxivity Contrast Agents in Peripheral MRA Jeffrey H. Maki1, George R. Oliveira1, Gregory J. Wilson2 1 - University of Washington Dept of Radiology, Seattle, WA. 2 - Philips Medical Systems, Cleveland, OH. Purpose Evaluate the utility of standard dose (~0.1 mmol/kg) high relaxivity (SDHR) gadolinium contrast for peripheral MRA (pMRA), and compare this to high dose high relaxivity (~0.2 mmol/kg) (HDHR) and high dose conventional relaxivity (HDCR) gadolinium contrast. Methods 60 patients undergoing 3 station moving table pMRA for suspected peripheral vascular occlusive disease received one of three contrast agents/doses (20 each): 17 cc gadobenate dimeglumine (MultiHance; Bracco Diagnostics), 34 cc gadobenate dimeglumine, or 34 cc gadoteridol (ProHance; Bracco Diagnostics). Imaging was performed on a 1.5T system (Gyroscan NT, Philips Medical Systems) using a prototype 18 channel coil and a SNR/timing optimized pulse sequence designed to maximize SNR while avoiding venous contamination1. Upper/middle scan times ranged from 5-12.5 sec, with lower station 45-60 sec. Arterial contrast ratios (CR: arterial SI minus muscle SI divided by muscle SI) were calculated for each station and corrected for differences in TR/TE. Subjective image quality and venous enhancement were evaluated. Statisitical analysis was performed using the Student t-test and Mann Whitney U tests. Results All studies were diagnostic and high quality. Contrast ratio was uniformly higher for HDHR contrast in all stations, but only significant in the lower station. Interestingly, upper station HDHR image quality was rated significantly worse than the others despite its higher CR. This trend has been noted in other studies, where high dose gadobenate dimeglumine appears to decrease MRA image quality/diagnostic efficacy2. There was no significant contrast ratio or image quality difference between SDHR and HDCR, however lower station venous enhancement was significantly less than the others for SDHR. Conclusion Excellent quality peripheral MRA can be performed with standard dose gadobenate dimeglumine, as overall quality is equivalent to or better than double dose conventional contrast. Increasing HR contrast dose improves CR, but may not improve image quality. Further investigation into this phenomenon is underway. References 1. Potthast et al. J Magn Reson Imag 29:1106-1115, 2009. 2. Schneider et al. J Magn Reson Imaging. 26(4):1020-32, 2007. 44 MR-Angioclub East Lansing 2009 4.3 Peripheral MRA With Continuous Table (CTM) Movement in Combination with High Temporal and Spatial Resolution TWISTMRA With 0.1 mmol/kg Gadobutrol at 3.0 T 1 Voth M , Haneder S1, Gutfleisch A1, Schoenberg SO1, Michaely HJ1 Institute of Clinical Radiology and Nuclear Medicine, University Medical Center Mannheim, Medical Faculty of Mannheim, University of Heidelberg, Germany 1 Purpose: To prove the concept of peripheral CTM-MRA in combination with high spatial and temporal resolution time-resolved TWIST-MRA in a single MR-exam at 3.0T with a single dose (0.1 mmol/kg) of gadobutrol in total. Methods: 22 consecutive patients (15m/7f, mean age 64 years) referred for peripheral MRA with clinical symptoms of peripheral arterial occlusive disease (PAOD) Fontaine stages II–IV underwent both CTM-MRA (TR 2.4ms/ TE 1.0ms/ flip angle 21°) of the run-off vessels and TWIST-MRA (TR 2.8ms/ TE 1.1ms/ flip angle 20°) of the calf station during a single MR-exam at 3.0T (Siemens Tim Trio). Spatial resolution of the CTM-MRA datasets was 1.2mm isotropic. The TWIST-MRA was acquired with 1.1x1.1x1.35mm³ and reconstructed to 1.1mm isotropic with a temporal resolution of 5.5 s. A total of 0.1 mmol/kg BW gadobutrol diluted 1:1 with saline was injected at a flow rate of 1.5 mL/s of which 0.07 mmol/kg was administered for the CTM-MRA and 0.03 mmol/kg for the TWIST-MRA. CTM-MRA run off datasets were qualitatively assessed using a four point scale (4excellent, 1-non-diagnostic) followed by TWIST-MRA datasets. Additional relevant findings only visible in the TWIST-MRA were documented. Results: All datasets could be evaluated with a total of 397 assessable segments. CTMMRA was diagnostic in 99% (393/397) with image quality judged as excellent in 54% (213/397), good in 42% (14/397), and moderate in 4% (14/397) respectively. Non diagnostic image quality was seen in 1% (4/397). TWIST-MRA was diagnostic in 100% (115/115) with good or excellent image quality. In 14 of 22 patients additional relevant findings were detected by TWIST-MRA. Conclusion: Single-dose gadobutrol CTM-MRA in combination with a high spatial and temporal resolution TWIST-MRA at 3.0 T is a reliable technique with good image quality. Despite the use of single dose contrast agent large field of view coverage and dynamic images can be acquired. Due to its robustness, this imaging approach of the vasculature has great potential for a broad clinical use. MR-Angioclub East Lansing 2009 45 4.4 Flow and Motion-Insensitive Unenhanced MR Angiography of the Peripheral Vascular System –A pilot study in the lower extremity John J. Sheehan 1,2, Ioannis Koktzoglou1, James C. Carr2, Eugene Dunkle1, Robert R. Edelman1 1 Department of Radiology, NorthShore University HealthSystem, 2650 Ridge Ave., Evanston, IL 2 Cardiovascular Imaging, Northwestern University, 737 N. Michigan Ave, Ste1600, Chicago, IL Introduction: Current limitations of unenhanced magnetic resonance angiography techniques include sensitivity to flow velocity, cardiac rhythm, and patient motion. We developed an alternative unenhanced method for peripheral magnetic resonance angiography (MRA) with the potential for robust performance over a wide range of physiological conditions. Materials and Methods: Flow-insensitive single shot (FISS) MRA acquires data with a specially modified single shot two-dimensional (2D) balanced steady-state free precession (bSSFP) pulse sequence. A key feature is the use of a quiescent inflow time period (QITP), coincident with systole, which is sandwiched between a saturation module and the bSSFP readout. The QITP provides the opportunity for maximal inflow of unsaturated arterial spins while ensuring suppression of venous signal. The combination of a saturation magnetization preparation with a single shot acquisition minimizes sensitivity to heart rate variations and arrhythmias. In order to test the clinical feasibility of the technique, we performed a pilot study of FISS MRA using contrast-enhanced MRA as the reference standard. A series of 4 healthy subjects (4 male, ages 28-45) and 8 patients (8 male, ages 56-90) with documented peripheral vascular disease were studied. Results: In all subjects, FISS MRA demonstrated the entire length of the peripheral vascular tree from aorta to pedal vessels. Considering CE-MRA as the standard of reference examination and excluding stented arterial segments, the sensitivity, specificity, PPV, and NPV values of FISS MRA for arterial narrowing greater than 50% or occlusion were 92.2%, 94.9%, 83.9% and 97.7% respectively. FISS MRA provided robust depiction of normal arterial anatomy and peripheral vascular disease, irrespective of disease severity, in scan times on the order of eight minutes for the entire peripheral vascular tree. In no subject was there substantial degradation of image quality due to bulk motion or variation in cardiac rhythm. Conclusion: FISS MRA is a fast, flow-insensitive and easy-to-use method for depicting the peripheral arteries. In a small group of patients, the unenhanced technique had excellent negative predictive value and image quality was consistent irrespective of the severity of PVD. Unlike subtraction-based unenhanced 3D MRA, the technique does not need to be tailored for each patient and initial results demonstrate reliable image quality for the pelvic vessels despite the presence of respiratory motion. 46 MR-Angioclub East Lansing 2009 4.5 A Comparison of Time-Resolved 3D CE-MRA with Peripheral Run-off CTA in the Calves CR Haider, JF Glockner, TJ Vrtiska, TA Macedo, EA Borisch, SJ Riederer MR Laboratory, Mayo Clinic, Rochester MN USA PURPOSE To perform a comparison of time-resolved 3D CE-MRA of the calves using Cartesian Acquisition with Projection-Reconstruction-like sampling (CAPR) with peripheral CTA. METHODS Nine patients who underwent clinically indicated peripheral run-off CTA were recruited for MRA within 48 hours of their CTA. CTA was performed using the standard protocol of our vascular CT clinical practice on a 64-detector row CT scanner (Sensation 64; Siemens Medical Solutions, Forschheim, Germany) resulting in 0.6 mm in-plane resolution with 2 mm thick axial sections and 1.2 mm increment. The CAPR acquisitions were performed on a 3.0 T MRI system (GE Healthcare, Milwaukee, WI.). 3D image sets with 1 mm isotropic spatial resolution were generated for a field of view of 40 cm S/I, 32 cm L/R, and 13.2 cm A/P using a 5 sec frame time and 20 sec temporal footprint [1, 2]. Using CTA as the reference, the MR data sets were evaluated with respect to depiction of any pathology, relative prominence of the luminal signal, and any added value of the time-resolved information. RESULTS The diagnostic image quality of the MRA results was rated very competitively vs. CTA (example, Figure 1). The CAPR sequence provided clear arterial frames in the case of rapid arterial to venous transit in patients with ulceration of the feet. In no MRA study was any reconstruction artifact observed that adversely affected image quality. (A) (B) (C) CONCLUSION These studies suggest that time-resolved MRA of the calves using CAPR can provide similar information to that generated using CTA. References: [1] Haider et al., MRM 60:749 (2008); [2] Haider et al., Radiology (in press). Figure 1. CTA vs. CAPR MRA. Vessel noted in CTA (A, arrow) is seen to fill retrograde in consecutive 5 sec frames in MRA (B-C). MR-Angioclub East Lansing 2009 47 4.6 3D Time-Resolved MR Angiography of Lower Extremities using Cartesian Interleaved Variable Density Sampling and HYPR Reconstruction 1a K. Wang, 2R. Busse, 1aY. Wu, 1aL. Keith, 2J. Holmes, 1a,bF. Korosec 1b Medical Physics, Radiology, University of Wisconsin-Madison, Madison, WI 2 Applied Science Lab, GE Healthcare, Madison, WI 1a Purpose: To simultaneously improve spatial and temporal resolution of 3D time-resolved MR angiography (TR-MRA) in lower extremities using Cartesian interleaved variable density sampling (VDS) [1] and HYPR [2-4]. Methods: Each time frame, a subset of k-space views are acquired with sampling density proportional to 1/kr as shown in Fig. 1 and described further in Ref. [1]. The sub-sampling patterns for a series of timeframes are designed to interleave, such that the data over N timeframes is fully sampled and HYPR methods in Ref. [3,4] can be used. Imaging parameters include: 0.94 x 0.94 x 1.5mm3 voxel size with matrix size of 512x282x72, 30 time frames with 5.8 sec/frame. Results: Fig. 2 shows coronal MIP images at arterial (a) and venous (b) phases. Compared with zero-filling, HYPR reduces the spatial blurring and increases the SNR, while preserving the temporal fidelity; View-sharing techniques yield early enhancement (thin arrow) and temporal blurring (thick arrow), as shown in Fig. 2(d). Conclusion: It is feasible to improve the spatial and temporal resolution in 3D TR-MRA using Cartesian interleaved variable density sampling and HYPR reconstruction. References: [1] Busse et al., ISMRM 2009, p4534. [2] Mistretta et al. MRM 55:30 (2006). [3] Wang et al. ISMRM 2009, p3884 [4] Busse et al., ISMRM 2009, p2834 in B being slightly different. After a short pause, a conventional high spatial resolution MRA (0.1 mmol/kg Gadovist®) was acquired using a 3D spoiled gradient-echo sequence 48 MR-Angioclub East Lansing 2009 4.7 3D Noncontrast MRA Using FSD-Prepared Balanced SSFP 1,2 1 3 1,2 3 1 1,2 Z. Fan , J. Sheehan , X. Bi , T. J. Carroll , R. Jerecic , J. Carr , and D. Li Radiology, 2Biomedical Engineering, Northwestern University, Chicago, IL, USA; 3 Siemens Medical Solutions USA, Inc., Chicago, IL, USA 1 Purpose: To develop a 3D noncontrast MRA (NC-MRA) method for peripheral arterial system using flow-sensitive dephasing (FSD)-prepared balanced SSFP (bSSFP). Methods: The FSD preparative module consists of a 90ox-180oy-90o-x pulse series and symmetric gradients around the 1800 pulse, which imparts flow sensitization measured by the first-order gradient moment (m1) [1]. An optimal m1 can selectively suppress the fastflow arterial blood signal during systole while having little effect on the slow-flow venous blood and static tissues. Subtraction between a bright-artery scan using bSSFP and a dark-artery scan using FSD-bSSFP will result in an artery-only dataset. The new technique was tested in multiple peripheral arterial territories of healthy volunteers and patients at 1.5-T. Contrast-enhanced (CE) MRA was performed in patient studies for reference. Parameters: 3D coronal acquisition, FSD gradients applied in the readout direction, isotropic high spatial-resolution (0.8 -1.2 mm), TE/TR =1.5-1.9/ 3.1-3.8 ms, flip angle = 70-90o, centric phase encoding, GRAPPA factor =2, m1 range 17-87 (leg), 58-156 (hand), 195-390 (foot) mT.ms2/m. Results: Proof of principle was obtained from the lower extremities and hands (Fig. 1-3). The value of m1 was shown to be critical for suppressing blood signal and achieving excellent MRA quality. Comparable depiction of stenosis/occlusion and superior visualization of patent segments were achieved using NC-MRA vs. CE-MRA. Conclusion: The feasibility of this NC-MRA approach has been successively demonstrated. Systematic optimization of m1 is warranted for clinical applications. References: [1] Koktzoglou I, et al. JCMR 9:33 (2007). a a b Fig.1 Both CE-MRA (a) and NC-MRA (b) demonstrate significant calf artery occlusion in a patient with PAD. The large collateral vessels are better depicted by NC-MRA. However, image artifacts due to metallic clip (arrow) and field inhomogeneity (arrowhead) are also appreciable. b Fig.2 NC-MRA (b) demonstrated more arterial vessels (in red) and detail than CE-MRA (a) in a patient with Raynauds disease. Fig.3 NC-MRA of the feet in a healthy volunteer. MR-Angioclub East Lansing 2009 49 4.8 Non Contrast MRA of the Hand in Patients with Raynauds disease using Flow Sensitized Dephasing Prepared SSFP 1,2 1 2 2 2 J. C. Carr , J. J. Sheehan , Z. Fan , A. Davarpanah , and D. Li Cardiovascular Imaging, Northwestern Memorial Hospital, Chicago, IL, United States, 2Cardiovascular Imaging, Northwestern University, Chicago, IL, United States 1 Introduction: Raynauds disease is vasospastic disorder of the digital arteries. 3D contrast-enhanced (CE) MRA is increasingly utilized for patients with Raynauds. Safety concerns with contrast administration in patients with renal insufficiency have led to a renaissance of non-contrast MRA (NC-MRA). NC-MRA strategies employing 3D halfFourier FSE [1] or SSFP [2] have shown great promise but various challenges remain. The aim of this study was to retrospectively assess the diagnostic quality and accuracy of a new NC-MRA method for hand MRA based on flow-sensitized dephasing (FSD)prepared SSFP. Materials and Methods: The proposed NC-MRA method acquires a bright-artery scan using ECG-triggered SSFP and a dark-artery scan using ECG-triggered, FSD-prepared SSFP [3]. Subtraction of the two scans results in bright arteries and suppression of the background and veins. 8 patients with clinically established Raynauds disease who have been imaged at 1.5T (Avanto, Siemens) using a 16-element peripheral matrix coil and spine coils, were retrospectively reviewed with IRB approval. Phase-contrast flow imaging was first performed to derive the arterial flow peak time T. Each patient subsequently underwent a time resolved TWIST MRA (TR MRA) and a high resolution MRA (HR MRA) with Gadolinium-BOPTA. The FSD subtraction images along with the TR MRA and HR MRA were reviewed and compared. Diagnostic quality was assessed by giving a per vessel score for the 17 segments per hand (1, poor; 2, fair; 3, good; 4, excellent) and adding them together for each hand (full score: 36). The degree of stenosis for each vascular segment was characterized using a four-point scale (grade 0, normal; grade 1, luminal narrowing <50%; grade 2, luminal narrowing >50%; grade 3 occlusion). Results: The mean qualitative score for TR MRA, HR MRA and FSD were similar: 3.8, 3.4 and 3.5, respectively (t-test, P>0.05). Differences were more apparent when comparing the distal digital arteries. Of 90 possible arterial segments, 6 (7%) were not adequately depicted with all techniques because of severe venous overlay. FSD identified 95% of luminal narrowings ≤50% and ≥50% that were identified on contrast enhanced MRA. Noncontrast FSD identified all of the arterial occlusions identified on contrast enhanced MRA. The mean quantitative scores were similar for the degree of stenosis for TR MRA, HR MRA and FSD were: 1.9, 1.9 and 2.0, respectively (t-test, P>0.05). Conclusion: Hi-resolution non contrast FSD MRA of the hand in patients with Raynauds compares favorably with contrast enhanced time resolved TWIST and high resolution static MRA and in many cases demonstrates improved resolution and visualization of normal and vasospastic vessels. References: 1. Miyazaki M, et al. Radiology 227:890 (2003). 2. Stafford R, et al. MRM 59:430 (2008). 3. Koktzoglou I, et al. JCMR 9:33 (2007). 50 MR-Angioclub East Lansing 2009 4.9 Two-Station Time-Resolved CE-MRA of the Lower Legs CP Johnson, CR Haider, EA Borisch, RC Grimm, PJ Rossman, SJ Riederer Department of Radiology, Mayo Clinic, Rochester MN 55902 USA Purpose: Contrast-enhanced MRA of the calves with high spatiotemporal resolution has recently been demonstrated [1,2]. The goal of this work was to demonstrate the technical feasibility of acquiring time-resolved extended field-of-view arteriograms of the thighs and calves, nearly doubling the longitudinal coverage of the calf studies while maintaining similar quality. Methods: Acquisition was based on the CAPR method [3] on a GE 3.0T MRI system. Two stations covering the thighs and calves of a volunteer were imaged with identical imaging parameters using a stepping table approach. 8x 2D SENSE was achieved with two high-performance eight-channel receive arrays [4], one placed at each station. A 2 mL timing bolus was used to determine the bolus transit time to the calves. The diagnostic scan was then acquired using 1.0 mm isotropic resolution with a 5.0 second image update time. 18 mL of Multihance followed by a 20 mL flush were injected intravenously at 3 mL/sec. Prior to the time-resolved sequence, a twostation calibration scan was acquired. Results: A maximum intensity projection of select arterial frames overlapped using a weighted sum is shown in Figure 1, demonstrating good image quality. Conclusion: Time-resolved CE-MRA of the lower legs using CAPR may allow for extended field-ofview arteriograms with diagnostic quality comparable Figure 1: MIP of select arterial frames. to that achieved in single-station studies. References: [1] Haider CR, ISMRM 2009, #2734. [2] Wu Y, JMRI 2009, 29:917-23. [3] Haider CR, MRM 2008, 60:749-60. [4] Johnson CP, ISMRM 2008, #1079. MR-Angioclub East Lansing 2009 51 5.1 An Update on the Clinical Experience with Gadofosveset 1 2 1 Stephen Schmitz ; Edward Parsons ; Mark G. Hibberd 1 Lantheus Medical Imaging, N. Billerica, MA, USA; 2 Independent Consultant, Burlington, MA Purpose: The clinical use of gadofosveset trisodium is reviewed in light of over three years’ use in Europe and Canada, the recent US-FDA approval of gadofosveset trisodium, and the establishment of a new clinical registry in the United States. Methods: The post-marketing registry that samples the clinical experience of roughly 70,000 exposures of gadofosveset trisodium in Europe was reviewed. The adverse event profile was summarized and compared with the data found in clinical trials and scientific literature describing gadofosveset and gadolinium contrast agents in general. The physiochemical and pharmacokinetic properties of gadofosveset were reviewed in consideration of leading hypotheses on the mechanism of NSF, and a new clinical registry in the US was proposed. Results: To date, only one blood-pool contrast agent (gadofosveset) has been registered for use in MRA. The utility of gadofosveset in vascular MRI has been detailed both by well-controlled clinical trials and by investigator-initiated studies. The adverse event profile reported in post-market assessments is consistent with that reported in clinical trials. The properties and pharmacokinetics suggest a relatively low risk of NSF, and no cases of NSF have been associated with gadofosveset to date. A new prospective registry will assess ongoing adverse event risks. Conclusion: Over three years of experience has expanded knowledge of the safety and efficacy of gadofosveset since its introduction in Europe and Canada. This information should benefit the decision-making of clinicians considering its use. The impeding introduction of gadofosveset to the US market will bring many of these advances to a large new group of physicians and patients. 52 MR-Angioclub East Lansing 2009 5.2 A Re-analysis of MS-325 (gadofosveset trisodium) Clinical Trial Data in Support of US-FDA Approval 1 Edward Parsons , Neil Rofsky2, Gary Stevens3, Margaret Uprichard1, Kirsten Overoye-Chan1, Andrew Uprichard1. 1 EPIX Pharmaceuticals, Inc.; 2 3 Beth Israel Deaconess Medical Center, Harvard Medical School; Dynastat, Inc. Purpose: To verify the clinical efficacy of MS-325 (gadofosveset trisodium) in terms of sensitivity and specificity for detection of stenotic lesions in MRA, as demonstrated in phase 3 clinical trials [1,2], fulfilling current US-FDA regulatory requirements for efficacy. Methods: Image data collected in support of registration of gadofosveset was re-read according to a rigorous interpretation protocol. The protocol was developed in consultation with the FDA and included additional steps for specific training of blinded MRA readers, use of a consistent paradigm for image assessment, strict categorization of any image artifacts that precluded diagnostic assessment, and a conservative statistical analysis with precise conditions on unanimous gains in sensitivity and non-inferiority of specificity. Three readers assessed non-contrast and gadofosveset-enhanced MRA data from two trials of patients with suspected aorto-iliac arterial occlusions. Assessments were compared to DSA results as a gold standard. Results: The sensitivity endpoints were similar to those of the prior reads, with smaller gains in specificity. All three readers showed statistically significant (p<0.001) gains in sensitivity (+20.4%, +12.2%, +14.9%%) with at least non-inferior specificity (+0.4%, +7.9%, +0.2%). Significantly fewer contrast-enhanced MRAs were considered uninterpretable when compared with non-contrast images (average of 8.6% vs 1.4%). Conclusion: The efficacy of gadofosveset was affirmed, resulting in US-FDA marketing approval [3]. More standardized reading guidelines produced more consistent results. References: [1] Rapp et al. Radiology. 2005 Jul; 236(1):71-8. Radiology. 2005 Sep; 236(3):825-33. [3] Drug Approval [2]. Goyen et al. Package: Vasovist. http://www.accessdata.fda.gov/drugsatfda_docs/nda/2008/021711s000TOC.cfm MR-Angioclub East Lansing 2009 53 5.3 Gadofosveset Excretion into Human Breast Milk Malene S. Thomsen1, Walter Gössler2, Uwe Lang3 Manuel Aigner1, and Manuela Aschauer1 1 Division of Vascular and Interventional Radiology, Department of Radiology, Medical University of Graz, Auenbruggerplatz 9, 8036 Austria, 2 Institute of Chemistry, Analytical Chemistry, Karl-Franzens 3 University, Universitaets-platz 1, 8010 Graz, Austria, Department of Obstetrics & Gynaecology, Medical University of Graz Purpose - Gadolinium (Gd) based contrast agents are generally accepted as pharmaceutical agents with few adverse effects. However, for the contrast media gadopentetate (Magnevist®) it is common practice that nursing mothers interrupt breastfeeding, and discard the milk the first 24 - 48 hours after injection of gadopentetate [1] . The aim of this current investigation was to evaluate if this guideline can be applied as well when using the contrast agent gadofosveset (Vasovist®). Methods - The Gd-content of breast milk was monitored up to 51 hours after injecting gadofoveset to a lactating patient (indication: objective evaluation of pulmonary artery embolism and deep venous thrombosis to plan the cava filter implantation) by collecting the breast milk with intervals of 2-4 hours and analyzing samples for the Gd-content applying inductively coupled plasma mass spectrometry (ICP-MS) Results - A significant result found was that gadofosveset was excreted from the body with a half-life time of around 26 hours, which was twice that of gadopentetate [2,3] . That gadofosveset stays longer in the blood circulation can most likely be explained by the fact that gadofosveset reversibly bind to proteins[4] whereas gadopentetate is highly water soluble and has low binding to proteins. Conclusion - From the study it was discovered that gadofosveset stays in the body twice as long as gadopentetate. Therefore the guideline might not be directly applicable. We recommended no brest feeding in this case for 48 hr. References - 1) Lin & Brown, J. Magn. Reson. Imaging, 2007 25, 884-899 2) Schmiedl et al, Am. J. Roentgenol., 1990, 154, 1305-6 3) Rofsky et al, J. Magn. Reson. Imaging, 1993, 3, 131-132 4) Goyen, Vasc. Health Risk Manag., 2008, 4, 1-9 54 MR-Angioclub East Lansing 2009 5.4 Overview of Gd-BOPTA Phase III Trial for CERMA: What are the results, and how do we move forward? Thomas M. Grist, M.D. University of Wisconsin – Madison PURPOSE Contrast-enhanced magnetic resonance angiography (CE-MRA) is widely used for imaging atherosclerotic disease in the carotid, renal, and peripheral arteries yet few multicenter trials have been reported that document the safety and efficacy of modern CE-MRA techniques. Studies were recently initiated to assess the safety and efficacy of a single 0.1 mmol/kg bw dose of gadobenate dimeglumine (Gd-BOPTA; MultiHance) for MRA in these three vascular territories. The objective of this presentation is to review the results of a multi-center trial and highlight the potential impact of the salient findings on MRA practice in the future. MATERIALS AND METHODS A total of 825 patients from Europe (63%), North America (27%), and South America (10%) were evaluated. Imaging of the carotid (N=247), renal (N=291) and peripheral (N=287) vessels was performed. 2D TOF and CE MRA was performed, and compared to intra-arterial DSA as a reference standard. Assessment of MRA images was performed at an independent core imaging laboratory by 3 experienced, fully-blinded radiologists per study unaffiliated with any study center. DSA images were reviewed in a similar fashion by a single reader. McNemar’s test was used to compare sensitivity, specificity, and accuracy of TOF and CE-MRA for detection of clinically significant steno-occlusive disease (>50% or 60% for carotid). RESULTS AND DISCUSSION The diagnostic accuracy of CE-MRA was significantly greater than TOF-MRA for most vessel territories and readers (p<0.05). Likewise, inter-observer agreement for CE-MRA was greater than 2D-TOF MRA. However, lower sensitivity values than anticipated were observed for CE-MRA due to methodological issues encountered in the study. The results and potential impact of these findings on acceptance of CE-MRA techniques will be discussed. MR-Angioclub East Lansing 2009 55 5.5 Safety of Gadobenate dimeglumine (Gd-BOPTA) in Cardiovascular Imaging of Pediatric Patients Guenther Schneider, Hellmut Schürholz, Arno Bücker, Peter Fries Homburg University Hospital, Homburg/Saar, Germany Purpose: The advent of nephrogenic systemic fibrosis (NSF) has brought the safety of Gd-based contrast agents into sharp focus. Nowhere is safety of greater concern than among pediatric patients who frequently require multiple contrast-enhanced (CE) MR examinations over an extended period of time. Gadobenate dimeglumine (MultiHance) is a contrast agent that has proven extremely safe among adult subjects for a variety of indications. Due to its increased relaxivity and the stability of the complex it seems to be well suited for cardiovascular imaging in pediatric patients. The present retrospective analysis was performed to determine the safety of gadobenate dimeglumine in pediatric subjects referred for routine cardiovascular MRI. Material and Methods: Gadobenate dimeglumine is routinely used off-label with IRB approval at our center for this specific patient population because of its beneficial properties. A total number of 88 pediatric patients (134 studies; 0 years - 15 years) underwent CE-MRA studies with a gadobenate dimeglumine dose of 0.1 mmol / kg bodyweight. Since only in-patients were studied, monitoring for adverse events was performed up to at least 24 hours post injection. Laboratory measurements, vital signs and ECG determinations were made before and after CE-MRA. Results: No severe or serious adverse events were noted in our series. No significant changes of creatinine, bilirubin, vital signs or ECG parameters were noted. No patients exhibited symptoms of NSF even after repeat doses of gadobenate dimeglumine. Image quality was excellent especially in patients that underwent CE-MRA for evaluation of CHD. Conclusion: Gadobenate dimeglumine at a dose of 0.1 mmol/kg bodyweight is a safe and effective contrast agent for cardiovascular imaging of pediatric patients. Concerning the safety of Gd-based MR contrast agents, the lower required dose of gadobenate dimeglumine seems to be beneficial for vascular imaging. 56 MR-Angioclub East Lansing 2009 5.6 Retrospective 7 year Study of the Incidence of Nephrogenic Systemic Fibrosis in Patients Investigated with Gadolinium Contrast-Enhanced Renal Magnetic Resonance Angiography Collidge T A, Brown, M, Rao, A, Thomson P C, & Roditi G Glasgow Royal Infirmary, Alexandra Parade, Glasgow G31 2ER Purpose: There is now well documented association between the development of the condition Nephrogenic Systemic Fibrosis (NSF) and high dose gadolinium contrast MRI examinations in patients with severe renal failure. We set out to assess the incidence of NSF in patients with differing degrees of renal impairment investigated by contrastenhanced renal MRA. Methods & Materials: Patients from 1998 to 2005 assessed through retrospective analysis of electronic patient record (EPR). Patients excluded if follow-up < 90 days. Analysis included renal function (CKD stage) at time of scan, time to first follow-up, total follow-up time, number of follow-up episodes, EDTA diagnoses, diagnoses of NSF, reports of pruritus or other skin disorder, episodes of gadolinium-enhanced MRI and cumulative gadolinium doses. Results: Of 1200 patients who underwent gadolinium-enhanced renal MRA studies 500 had follow-up documented > 90 days with average follow-up of 1326 days. CKD eGFR MRA 5 <15 67 NSF 1 4 15-29.9 169 2 3 30-60 222 0 1&2 >60 31 0 Normal >90 11 0 (potential live renal donors) The 3 patients who developed NSF during the course of follow-up had estimated GFRs of <1, 16 & 18. One patient with initial CKD stage 4 at time of renal MRA developed NSF at a significantly later stage following a peripheral run-off MRA when clinically unwell in CKD stage 5. 5 other patients were noted at clinic follow-up to be complaining of pruritus but none of these was ever diagnosed as suffering from NSF. Conclusion: This study confirms that NSF is not seen in patients undergoing contrast- enhanced MRA with less severe degrees of renal failure than CKD stages 4 & 5 MR-Angioclub East Lansing 2009 57 5.7 Risk Factors for NSF: a Meta-analysis 1, Martin R. Prince 2, MD, PhD, Hong Lei Zhang1, MD, Giles H. Roditi3, MD 4 5 Tim Leiner , MD, Walter Kucharczyk , MD From Departments of Radiology at Weill Medical College of Cornell University1, Columbia College of Physicians and Surgeons2, New York, Glasgow Royal Infirmary, Scotland3, Maastricht University 4 5 Hospital, Maastricht, the Netherlands , and University of Toronto, Toronto, Canada Purpose: To investigate who can safely undergo Gd MRA with minimal risk of NSF. Method: Pubmed was searched for ‘Nephrogenic Systemic Fibrosis’ to identify papers with detailed clinical information from January, 2003 to June, 2009. Patient age, gender, race, type of Gd enhanced imaging, Gd type and dose, interval between Gd and NSF onset, GFR, dialysis, interval between Gd and dialysis, acuteness of renal failure, kidney transplant, serum phosphorus, acidosis, epoetin, pro-inflammatory events and NSF symptoms, were recorded and analyzed. Corresponding authors were contacted for corroboration and to provide as much missing data as possible. Results: 77 papers with detailed information on 290 patients were included in the final analysis. The gender distribution was approximately equally weighted. Very young (<10 yrs) and old (> 60 yrs) patients are at less risk. In 244 patients for whom MR exam history was investigated, 217 (89%) were noted to have GBCA injections prior to NSF symptom onset including use of gadodiamide (n = 167), Gadopentetate dimeglumine (n = 11), gadoversetamide (n = 5), multiple agents (n = 8) and unspecified/unknown (n = 26). All NSF patients had renal dysfunction with highest prevalence for dialysis and for GFR < 15 mL/min. Acute renal failure was an NSF risk factor with odds ratio of 13:1. In 182 cases for which data on GBCA dose were available or could be estimated, 90% (n = 163) received high dose. Other risk factors include pro-inflammatory events, epoetin, acidosis and hyperphosphatemia. Risk of NSF is small compared to the risk of iodinated contrast induced nephropathy and the risk death from NSF was small compared to the risk of death from allergic reactions. Conclusion: Order of magnitude reductions in NSF risk can be attained by 1) avoiding high dose, 2) avoiding nonionic linear chelates, 3) dialyzing within 24 hours of Gd administration for patients already on dialysis, 4) avoiding injecting acute renal failure patients while serum creatinine is rising. 58 MR-Angioclub East Lansing 2009 5.8 Extracellular matrix metabolism in organ-cultured skin from patients with end-stage renal disease: Response to gadolinium based MRI contrast agents James Varani, Marissa DaSilva, Monica O’Brien Deming, Kent Johnson and 1Richard Swartz; 1 Departments of Pathology and Medicine, University of Michigan, Ann Arbor, Michigan, 48109 Purpose: Nephrogenic systemic fibrosis (NSF) is a clinical syndrome linked with exposure in renal failure patients to gadolinium - based MRI contrast agents (GBCAs). The present study addresses potential patho-physiological mechanisms. Methods: Here we have examined human skin from normal subjects and individuals with end stage renal disease for response to GBCA stimulation in organ culture. Results: Omniscan, one of the clinically used GBCAs, had no effect on type I procollagen production, but increased levels of both matrix metalloproteinase-1 (MMP-1) and tissue inhibitor of metalloproteinases-1 (TIMP-1). The level of TIMP-1 was significantly higher than the level of MMP-1 and there was no detectable collagenolytic activity. Qualitatively, there were no differences in the responses of skin from renal failure patients as compared to controls. However, basal responses were higher in skin from subjects in renal failure as compared to control. Conclusion: These data suggest that GBCA exposure does not directly stimulate collagen production but, rather, modulates an enzyme-inhibitor system responsible for regulation of collagen turn-over in the skin. We speculate that heightened responses in subjects with end-stage renal disease may reflect elevated basal activity. MR-Angioclub East Lansing 2009 59 5.9 Manganese Based Biodegradable Macromolecular MRI Contrast Agents for Cardiovascular Imaging 1 1 1 1,2 Zhen Ye , Eun-Kee Jeong , Dennis L. Parker , Zheng-Rong Lu Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, Salt Lake City, Utah; 2 Department of Biomedical Engineering, Case Western Reserve University, Cleveland, Ohio 1 Purpose: This study was intended to develop Mn(II) based biodegradable macromolecular contrast agents for cardiovascular MRI. Methods: The polymeric contrast agents (CAs), Mn(II)-DTPA cystamine copolymers (MDCC) and Mn(II)-EDTA cystamine copolymers (MECC), were synthesized and characterized(1). Their in vivo contrast enhanced MR cardiovascular imaging was evaluated in mice on a Siemens 3T MRI scanner with MnCl2 as a control. Results: The number average molecular weight of MDCC and MECC was 30.5 and 60.8 KDa, respectively. The T1 relaxivity of MDCC and MECC was 4.7 and 6.41 mM-1s-1 at 3T, respectively, in the same range of that of MnCl2 (5.2 mM-1s-1)[2]. Contrast enhancement was observed in the vasculature with MDCC and MECC in the initial period post injection. Significant enhancement in the myocardium was also observed for MECC and MnCl2, not for MDCC. The difference in myocardium enhancement between MDCC and MECC might be attributed to the different coordination chemistry and the stability of two contrast agents[3]. - OOC N O O N N Mn2+ COO - COO N H N S S n MDCC -OOC -OOC Mn2+ O N H N H O N N H H N S S n pre MECC MnCl MDCC MECC 2 (30 minutes after i.v injection) Conclusion: The Mn(II) based biodegradable macromolecular contrast agents MDCC and MECC are promising for contrast enhanced cardiovascular imaging with MRI. Reference: [1] Intl. J. Nanomed. 2007, 2, 191-9. [2] Chem. Rev. 1987.87.901-927; [3] Cur.Pharm.Biotech, 2004, 5, 539-549 60 MR-Angioclub East Lansing 2009 6.1 Non-invasive Trans-Stenotic Pressure Measurements with 3D Phase Contrast MRA: Validation against Endovascular Pressure Measurements in Swine 1, 3 T Bley, 2K Johnson, 1C François, 1,2S Reeder, 1M Schiebler, 2O Wieben, 1T Grist 1 2 Departments of Radiology and Medical Physics, UW Madison, WI, USA Department of Radiology, University Medical Center Hamburg-Eppendorf, Germany 3 Purpose To evaluate trans-stenotic pressure gradients (TSPG) in renal artery stenosis (RAS) noninvasively utilizing a 3D phase contrast acquisition with vastly under sampled isotropic projection reconstruction (PC-VIPR) MRA in a porcine study. Methods Respiratory gated PC-VIPR MRA (dual echo, 18,000 projection angles, 10˚ flip, TR/TE (first echo) = 11.4/3.7 msec, BW = ±62.5kHz, imaging volume: 260x260x160mm3, acquired isotropic spatial resolution: 1.0x1.0x1.0mm3, venc = 150cm/s) of renal arteries with surgically created RAS in 12 swine was performed on a 1.5T clinical system (Signa HDx, GE Healthcare, Waukesha, WI). NCE MRA and TSPG were calculated directly from the magnitude and velocity measurements using the Navier-Stokes equation. Endovascular pressure measurements were used as the gold standard for pressure gradient quantification, performed under DSA guidance. Results In 19 renal arteries, the TSPG analysis demonstrated excellent correlation between the non-invasive TSPG utilizing PC-VIPR and endovascular pressure measurements (average stenosis = 62%, r = 0.977; 95% CI: 0.931, 0.998; p < 0.001). In 5 arteries with severe RAS (mean 86%), the residual lumen within the stenosis was so small that TSPG could not be determined using PC-VIPR and they were excluded from the statistical analysis. However, the angiographic reformats from those data readily revealed severe, hemodynamically significant RAS. Conclusion Noninvasive assessment of hemodynamic significance of RAS in swine was feasible utilizing non contrast material enhanced PC-VIPR MRA. Excellent correlation between TSPG measurements by PC-VIPR and endovascular guide wires was found. PC-VIPR has the potential to become a major advance in the noninvasive evaluation of RAS and, as a result, in the management of patients with renal hypertension. MR-Angioclub East Lansing 2009 61 6.2 Analysis of aortic hemodynamics after treatment for coarctation using flow-sensitive 4D-MRT at 3T A. Frydrychowicz1, D. Hirtler2, R. Arnold2, A. Berger1, A.F. Stalder1, J. Bock1, M. Langer1, J. Hennig1, M. Markl1 1 Department of Diagnostic Radiology, Medical Physics, University Hospital Freiburg, Germany, 2 3 Department of Pediatric Cardiology, University Hospital Freiburg, Germany Department of Cardiology, University Hospital Freiburg, Germany\ Purpose: To evaluate the hemo-dynamic alterations in aortic blood flow after coarctation repair by flow-sensitive, Fig. 1: Vortical flow patterns (white time-resolved 3D MRI at arrows) observed in 15/24 patients 3T and to compare after treatmen of aortic coarctation. findings to acquisitions in This finding could be observed volunteers. irrespective of the treatment Methods: Flow-sensitive methodology. This patient also 4D MR was performed in shows flow acceleration over a 28 patients after mild re-stenosis (openwhite coarctation repair arrow). (16.5±8.0years, range 436) and 19 volunteers (35.3±18.1 years, range 20-75) from a previous analysis were included as a reference [3]. Results: The operative site showed a relative restenosis with a diameter of 13.3 ± 4.0mm (range 6-24) and a post-stenotic dilatation of 18.8 ± 6.5mm (range 7-43). Next to accelerated flow in all patients, there were additional helices, an Fig. 2: Different stages of aneurysm development were increased rate of vortices detectable. (A) represents flow patterns in a 10yo boy and most noticeably, in before treatment of an otherwise not symptomatic the majority of all coarctation. Clearly, the elongated aortic arch, the patients unexpected vortical flow acceleration (white arrowheads) before vortices in the the stenosis (white arrow, moderate flow acceleration supraaortics that might during late systole) can be apreciated. In (B), a attribute to intimal hypoplastic arch (white arrow) gives rise to a flow thickening were found. acceleration taht enters a highly vortical flow through a Quantitative analysis small aneurysm that distributes greatly helical flow in showed significantly the Dao (yellow arrows). elevated wall shear stress levels and decreased oscillatory shear indices in the patients that hint towards an influence concerning arterial remodeling. Conclusion: Flow-sensitive MRI revealed marked changes in the hemodynamics after CoA repair. Follow-up examinations have to clarify, whether a predictive value can be attributed to hemodynamics or derived vessel wall parameters. References: 1. Oliver JM J Am Coll Cardiol 2004, 44:1641-1647. 2. Wigström L, et al. MRM 1999;4:793-799, 3. Frydrychowicz A, ISMRM 2008 62 MR-Angioclub East Lansing 2009 6.3 Flow assessment of arterial dissections using 3D radial phase contrast MR angiography Christopher J François, Kevin M Johnson, Benjamin Landgraf, Mark L Schiebler, Scott B Reeder, Thomas M Grist, Oliver Wieben University of Wisconsin, Madison, WI, USA Purpose: The ability to predict the progression and complications of dissections (eg. aneursymal dilatation and rupture) is a limitation of current diagnostic methods. Differences in the flow patterns in the true and false lumina may be important in the outcomes of dissections1-3. Our aim was to develop a 3D phase contrast (PC) radial pulse sequence that can be peformed in less than 10 minutes to assess the hemodynamics of dissections, as a potential prognosticator of dissection A evolution. Methods: 3D PC VIPR was performed in patients with arterial dissections. Parameters for PC VIPR were: ±62.5 kHz receiver bandwidth, 1.00-1.25mm3 isotropic spatial resolution, 8-10 min of free breathing with 50% respiratory gating efficiency, imaging volume: 32x32x16 cm3, VENC of 80-100 cm/s, retrospective cardiac gating with a temporal filter for radial acquisitions. PC VIPR data were acquired after obtaining patient consent according to our IRB protocol. Datasets were analyzed using EnSight (CEI, Inc., Cary, NC). Results: Flow patterns in true (arrows) and false (open arrows) lumina were laminar and non-laminar, respectively. In a patient with celiac and superior mesenteric artery dissections (A), the flow in the false lumina was vortical. In a B patient with chronic descending thoracic aortic dissection (B), flow in the false lumen was very slow and turbulent. Conclusion: Using PC VIPR it is possible to detect differences in flow patterns between true and false lumina of arterial dissections, which should permit a more comprehensive quantitative hemodynamic assessment and could have prognostic implications in addition to improving our understanding of the underlying pathophysiology. References: 1. Markl M, et al. JCAT 2004;28:459. 2. Strotzer M, et al. Acta Radiol 2000;41:594. 3. Mohri M, et al. Clin Cardiol 1985;8:225. MR-Angioclub East Lansing 2009 63 6.4 High Temporal and High Spatial Resolution Perfusion Imaging of Hepatocellular Carcinoma in the Liver Scott B. Reeder, MD, PhD; Ethan K. Brodsky, PhD; Eric Bultman, BS; William Schelman, MD, PhD; Yin Huang, PhD; Sean F Fain, PhD; Walter F. Block, PhD Departments of Radiology, Medical Physics, Biomedical Engineering and Medicine University of Wisconsin, Madison, WI, USA Introduction: The field of cancer therapeutics has been shifting from traditional “cytotoxic” agents to targeted anti-angiogenic therapies. This trend requires biomarkers of early tumor response. Here we aim to develop high spatial and temporal resolution imaging methods for quantitative perfusion imaging as a biomarker of tumor response. Methods: Patients with known hepatocellular carcinomas (HCC) were imaged on a 3.0T scanner (MR750, GE Healthcare) with a 32 channel abdominal coil, after single dose injection of gadobenate dimeglumine (Bracco, Princeton, NJ). A time-resolved 3D radial sequence collected whole abdomen data using one interleaved sub-frame/sec, reconstructed in real-time1, allowing fluoroscopic monitoring of the contrast bolus. Real time bolus tracking permits timing of breath-holding during 1) contrast arrival during the arterial phase, 2) portal venous phase (50-70s), and 3) delayed phase from 1:40-2:00 minutes. Using temporal filtering through density compensation filters2, time-resolved image volumes at full spatial resolution (1.6x1.6x1.6mm3) were reconstructed with an effective temporal resolution of 8s. Using the high temporal resolution images, quantitative perfusion modeling can be performed using a dual input model of the hepatic blood flow that includes the portal vein and hepatic arterial input3. Results and Discussion: Figure 1 shows images of a 1.5cm HCC, identified on surveillance MRI. Arterial phase shows brisk enhancement and rapid washout during the portal venous phase, highly characteristic of HCC. Spatial resolution is 1.6mm isotropic with 8s temporal resolution. High spatial and temporal resolution will allow clinicians to accurately capture contrast uptake in the tumor contrast kinetics, the portal vein and hepatic artery, necessary for dual input modeling of perfusion to the HCC3. Figure 1: Pre-contrast, arterial phase, and portal venous phase images through a 1.5cm HCC (arrow). Brisk arterial enhancement and rapid washout is highly characteristic of HCC. Spatial resolution is 1.6mm isotropic with 8s temporal resolution. References: 1. Brodsky et al, MRM, 2006 2. Barger et al, MRM, 2002 3. Materne et al, MRM, 2002 64 MR-Angioclub East Lansing 2009 6.5 Stack of Stars 4D Phase Contrast Velocimetry of the Circle of Willis Steven Kecskemeti, Kevin Johnson, Yijing Wu,Warren Chang, Charles Mistretta, Patrick Turski Departments of Medical Physics and Radiology, University of Wisconsin, Madison, WI INTRODUCTION: In addition to angiograms, cardiac gated phase contrast (PC) MR velocimetry provides hemodynamic information such as pulsatility, flow streamlines, relative pressure, and estimates of wall shear stress. In certain cases, such as aneurysms in the circle of Willis (COW), high resolution over a moderate excitation slab (40mm) can be achieved using a stack of stars hybrid PR acquisition. METHODS AND RESULTS: A cardiac gated (both retrospective and prospectively) gated PC stack of stars (SOS) sequence was developed with radial readout in the xyplane and traditional phase encoding in the z-direction. A bitreversed projection ordering allows for retrospective viewsharing in multiples of the minimum temporal width of 4TR for four point velocity encoding. Exams have been performed on a 3T clinical scanner with 8-channel head coil. Parameters were FOV: 220x220x50mm, resolution:0.43x0.43x1.0mm, reconstructed voxel size: 0.43x0.43x0.43mm, TR/TE = 9.3/4.1ms, tip angle = 20, BW = 62.5kHz, scantime 7:03+30s calibration scan. Full MIPs have been shown for the axial (a) and coronal(d) views, while limited MIPS of sagittal are displayed to distinguish left (c) and right (d) hemispheres. We conclude that 4D cardiac gated PC SOS a practical method to study flow normal and pathological flow conditions within the Circle of Willis. Mean Flow 2.8 flow [ml/s] 2.6 2.4 2.2 High Resolution Perfusion Weighted Imaging 2 0 200 400 600 800 1000 time [ms] MR-Angioclub East Lansing 2009 65 6.6 High Resolution Perfusion Weighted Imaging Meng Li, MS and E. Mark Haacke, PhD Wayne State University, Detroit, MI, USA Purpose: Perfusion-weighted imaging (PWI) using dynamic susceptibility contrast (DSC) is a useful tool to evaluate various diseases of the brain. But current low resolution PWI is unable to easily show differentiation between gray matter and white matter. We plan to use susceptibility weighted imaging (SWI) and MR angiography (MRA) to remove major vessel information from a high resolution PWI approach to better reveal gray matter perfusion without the confounding major vessel effects. Methods: HR PWI images of 1 x 1 x 4 mm3 were acquired on a 1.5T Siemens Sonata with an 8-channel head coil. The imaging parameters TR/TE/FA/Resolution/BW=2200ms/98ms/60°/1x1x4mm3/752Hz/pixel, used the were: acquisition matrix was 256x256 (interpolated to 512x512). Both MRA and SWI data were acquired. MRA was obtained pre-contrast with TE/TR/FA = 7ms/37ms/25°, SWI was acquired pre and post-contrast with TE/TR/FA = 40ms/49ms/20°. Results: High resolution PWI was successfully obtained with parallel imaging. Compared to the usual 2mm in-plane resolution, the following advantages were observed on CBF and CBV maps: the blurring of blood vessels was minimal, fine details of blood vessels were observed, gray matter was clearly separated from blood vessels, and small vessels like the medullary veins could be seen. Using the HR PWI data it was possible to remove the blurring associated with the major blood vessels. This is accomplished by using both the MRA and SWI data. The major vessels were then removed from the HR PWI CBV map in an attempt to create a segmented vessel free image. This made it possible to focus on just gray matter or white matter and made it easier to filter the gray matter data without interference from major vessels. Conclusion: The separation of gray matter and white matter in PWI can be achieved by increasing the in-plane resolution of conventional PWI data to 1mm. Combination of high resolution maps of PWI, SWI and MRA can provide additional information about macro and micro circulations and vasculature of the brain. 66 MR-Angioclub East Lansing 2009 6.7 Accelerated velocity imaging using compressed sensing 1 1 2 1 1 L. Marinelli , K. Khare , K. F. King , R. Darrow , T. K. F. Foo1, and C. J. Hardy 1 2 GE Global Research, Niskayuna, NY, GE Healthcare, Waukesha, WI Purpose: Accurate measurement of blood velocity in complex flows can improve the diagnosis and characterization of a variety of cardiovascular diseases. We have developed a 2D Fourier velocity encoding (FVE) M-mode MRI pulse sequence to probe multi-dimensional velocity distributions. Unlike in conventional MRI, parallel imaging cannot be used to shorten FVE scan time. We have instead developed a compressed sensing (CS) approach to accelerate 2D FVE, which exploits the sparsity of the blood velocity distribution. Methods: Following localization of the mitral valve, 2D FVE M-mode MRI was performed with a VENC of 64 cm/s. A 2-cm pencil of spins was excited and 16 velocity-phaseencoding steps were applied in each velocity-encoding direction. The 16x16 velocity encodings were undersampled by various factors and we compared uniform random, Gaussian random, and Poisson disk distributions. Results and Conclusions: Figure 1 shows a typical temporal evolution of the blood velocity distribution near the mitral valve. Frame (a) was acquired 20ms after the QRS complex and in frame (c) we note that there was some flow towards the aortic valve. During diastole, the aortic valve closes and the mitral valve opens; flow near the mitral valve becomes more directional towards the valve (frame (e)) and then relaxes back to zero velocity sweeping an arc in velocity space (frame (f) and (a)). This is the only technique that provides multi-dimensional velocity distribution data that may have impact in the assessment of valvular disease and regional intravascular wall pressure. vx 1/32 a vz 3/32 b 8/32 c 18/32 11/32 d e 21/32 g f Figure 1. (a)-(f)Time evolution of the 2D blood velocity distribution near the mitral valve (green arrow in (g)). vx (resp. vz) is the velocity component parallel (perpendicular) to the pencil beam. (g) Left h ventricular outflow tract view with M-mode pencil prescribed through the mitral valve. (h) 2D FVE pulse sequence diagram. MR-Angioclub East Lansing 2009 67 6.8 Wall Shear Stress in Normal and Atherosclerotic Carotid Arteries M. Markl, S. Bauer, J. Bock, A. F. Stalder, A. Frydrychowicz, A. Harloff 1 Diagnostic Radiology, Medical Physics, University Hospital Freiburg, Germany Introduction: Hemodynamic conditions in the carotid bifurcation resulting in low wall shear stress (WSS) and high oscillatory shear index (OSI) can predict plaque development1,2. The normal distribution of segmental WSS and OSI was evaluated in 32 normal volunteers and compared to findings in 6 patients with moderate internal carotid artery (ICA) stenosis < 55%. Methods: Flow-sensitive 4D-MRI3,4 (a=15°, venc=150cm/s,TRes=45.6ms, 1.1x0.9x1.4mm³) was performed to estimate time-averaged absolute WSSmag and OSI in 7 analysis planes (figure 1)5. To identify areas at risk for plaque development segments representing the individual upper 15% of OSI and lower 15% WSSmag were determined. Results: 64 normal carotid bifurcations (figure 2) revealed high inter-individual consistency and a high incidence of low WSSmag and high OSI in segments corresponding to the proximal ICA bulb. Patients showed a more heterogeneous distribution and a spatial relocation of critical wall parameters to proximal regions distal to the ICA stenosis. Discussion: Atherogenic low WSSmag and high OSI in the normal ICA bulb may explain why ICA plaques often develop at this site. The presence of ICA stenosis can alter wall parameter distributions which may help to predict the direction and extent of progression of the disease. Fig. 1: Assessment of wall parameters in 7 analysis planes (left) and extracted systolic velocity profile and wall shear stress vectors (WSS) for analysis plane 3 in the ICA bulb (right). Fig. 2: Distribution of critical wall parameters for n=64 normal carotid arteries and n=6 patients with moderate ICA stenoses. In normal controls critical wall parameters were predominately located in the posterior ICA bulb while patients showed critical WSS in more distal segments. References: 1. Cheng C et al. Circulation. 2006;113:2744-2753; 2. Lee SW et al. Stroke. 2008;39:2341-7; 3. Harloff A, et al. MRM 2009, 61:65-74; 4. Markl M, et al. JMRI 2007, 25:824-831; 5. Stalder AF et al MRM. 2008;60:1218-1231. 68 MR-Angioclub East Lansing 2009 6.9 MR imaging and significance of flow reversal and carotid atherosclerosis: Initial results 3 J. Scott McNally , Seong-Eun Kim1,3, John Roberts1,3, Gerald S. Treiman2,4, Dennis L. Parker1,3 1 Utah Center for Advanced Imaging Research, 2VA Salt Lake City Health Care System, 3Radiology, 4 Surgery, University of Utah, Salt Lake City, Utah, 84108 Introduction Reports suggest that atherosclerotic plaque at the carotid bifurcation correlates with areas of flow reversal and low wall shear stress. MRI can be used to delineate plaque components including ulceration, erosion, and hemorrhage, which are thought to correlate with symptoms. Using MRI, our goal is to determine if areas of flow reversal correlate with these markers of plaque vulnerability and plaque location at the carotid bifurcation. Methods VA subjects scheduled for carotid endarterectomy are being scanned with a 3T MRI using the following sequences: 2D T1w (0.5 x 0.5 x 2mm3) and 2D TSE to measure plaque location, ulceration and erosion, 3D MPRAGE (0.5 X 0.5 x 1mm3) to measure hemorrhage, 3D TOF (0.3 x 0.6 x 0.6mm3) and 2D Phase Contrast (0.8 x 0.8 x 3.0mm3) with dual VENC to measure flow reversal. After tracing the regions of interest, areas of these components are calculated. Results and Discussion Results obtained to date on subjects with atherosclerosis (16 diseased and 7 normal carotid arteries) have shown: 1) Flow reversal in normal carotid arteries occurs along the proximal internal carotid artery (ICA) and downstream of plaques. 2) Plaque area is largest along the outer wall of the ICA. 3) Ulceration, erosion and hemorrhage occur preferentially in areas exposed to flow reversal. Conclusion Initial results suggest that flow reversal correlates with plaque location and markers of vulnerability. This study may help determine the significance of flow reversal in the progression of atherosclerosis. MR-Angioclub East Lansing 2009 69 7.1 4D-MRA in combination with arterial spin labelling for selective and functional information in patients with AVMs W. A. Willinek1,G.M. Kukuk1, D.R. Hadizadeh1, J. Gieseke1, 2, J. Bergener1, G. Beck2, L. Geerts2, P. Mürtz1, A. Boström3, H. Urbach1, J Schramm3, H. H. Schild1 1 Dpt. of Radiology, University of Bonn, Germany, 2 Philips Healthcare, Best, Netherlands, 3 Dpt. of Neurosurgery, University of Bonn, Germany Purpose: Arterial spin labelling with selective labelling pulses is a promising method providing selective and functional information regarding brain perfusion territories, regional cerebral vascular supply and functional collateral circulation. The purpose of this study was to prospectively evaluate 4D-MRA in combination with selective arterial spin labelling for perioperative assessment of cerebral AVMs. Methods: In a prospective intraindividual comparative study 10 patients (6 female, 4 male; mean age 35.8 years ± 12.2; range 20-58 years) diagnosed with symptomatic cerebral AVMs underwent pre- and postoperative 4D-MRA, regional brain perfusion imaging using selective arterial spin labelling and DSA. Institutional ethics committee approval and written informed consent were obtained. 4D-MRA was performed using CENTRA keyhole in combination with view sharing yielding a temporal resolution of 572 msec, whole brain coverage and an isotropic voxel size of 1.1 x 1.1 x 1.1 mm³. Selective arterial spin labelling was performed using the PULSAR labelling sequence for selective labelling of both carotid arteries and the vertebrobasilar complex. All images were pre- and postoperatively assessed by two radiologists in consensus regarding technical success rate, preoperative assessment (Spetzler-Martin classification, identification of arterial feeders, existence of anatomic variants / functional crossfilling) and completeness of resection. In all cases DSA served as the standard of reference. Results: 4D-MRA was successfully performed in 20/20 exams and enabled the same Spetzler-Martin classification as DSA in all cases (100 %). 11/13 (85 %) feeding arteries were identified by 4D-MRA. Selective arterial spin labelling was successful in 16/20 (80 %) exams. Selective arterial spin labelling provided additional functional or anatomic information in 2/16 exams and enabled the diagnosis of a cross-filling feeding artery that was not identified by 4D-MRA but by DSA, thus improving the sensitivity of MRI in identification of arterial feeders from 11/13 (85 %) to 12/13 (92 %). Postoperative assessment confirmed complete resection of all AVMs in 100 % yielding a 100 % concordance between 4D-MRA in combination with selective arterial spin labelling and DSA. Conclusion: 4D-MRA in combination with selective arterial spin labelling is a promising tool for pre- and postoperative assessment of cerebral AVMs providing functional information that so far has been gained only with selective DSA. References: [1] Golay X, Petersen ET, Hui F. Pulsed star labeling of arterial regions (PULSAR): a robust regional perfusion technique for high field imaging. Magn Reson Med 2005;53:1521. [2] Hendrikse J, van der Grond J, Hanzhang L et al. Flow territory mapping of the cerebral arteries with regional perfusion MRI. Stroke 2004;35:882-887. [3] Lim CC, Petersen ET, Ng I et al. MR regional perfusion imaging: visualizing functional collateral circulation. AJNR 2007;28:447-448. [4] Willinek WA, Hadizadeh DR, von Falkenhausen M. 4D time-resolved MR angiography with keyhole (4D-TRAK): more then 60 times accelerated MRA using a combination of CENTRA, keyhole and SENSE at 3.0T. JMRI 2008;27:1455-1460. [5] Hadizadeh DR, von Falkenhausen M, Gieseke J et al. Cerebral arteriovenous malformation: Spetzler-Martin classification at subsecond-temporal-resolution fourdimensional MR angiography compared with that at DSA. Radiology 2008;246:205-213. 70 MR-Angioclub East Lansing 2009 7.2 Intraindividual comparision between multislice CT and 4 D TWIST MRA in the assessment of residual cerebral arteriovenous malformations – a prospective study protocol Authors: M. Essig, M. Voth, A. Zabel-Du-Bois, L. Schuster Institution: Department of Radiology, German Cancer Research Center, Heidelberg, Germany Rationale and Objectives Small or residual cerebral arteriovenous malformations (AVMs) are hard to visualise with MR angiographic techniques. Although MRA techniques like TOF or contrast enhanced MRA allow high resolution they are not able to visualize small or slow flowing vessel segments. However, these small vascular structures and their intra cranial course are important for radiotherapy planning and follow-up assessments. Therefore MS-CTA is used to assess complete obliteration if MRA proved negative. The aim of the presented study protocol was now to assess the validity of a time resolved 3D contrast enhanced MRA technique in direct comparision with CTA. Methods In an ongoing dual centric prospective study protocol we used a combination of high temporal resolution (250ms) and high spatial resolution (1x1x1 mm) CE MRA using a standard dose (0,1mmol) of the macrocyclic Gadobutrol (Gadovist®, Bayer, Berlin) for a TWIST (Time-Resolved Imaging with Stochastic Trajectories) acquisitions. The new 4 D MRA technique was intraindividually compared with MS CTA to assess the visibility of residual small AVM compartments and the presence of arteriovenous shunting. The imaging data were evaluated in a qualitative matter by two independent readers with special attention to the morphologic features of the malformation and the influence of the dynamic MRA studies. Results So far we were able to include 12 patients with no residual AVM components on conventional TOF-MRA examinations (inclusion criteria). In10 out of these patients residual AVM components could be visualized on the dynamic and high resolution contrast enhanced MRA sequences. The MS CTA was able to visualize residual vessel components in only 8 patients. In two patients massive artifacts from previous embolisation hindered the visualisation. In two patients no vascular components were seen on both CTA and contrast enhanced MRA. In the MRA studies the hemodynamic information from the TWIST technique was found to be very helpful in the differentiation between physiologic and pathologic vessels. Conclusion In conclusion these preliminary results could prove that high temporal and spatial resolution 4D MRA is able to assess even small residual AVM compartments at a high sensitivity. Based on the initial patient studies, Gadobutrol with its unique high molar concentration seems to be the ideal contrasting agent in this study concept. The high concentration allowed for a small bolus which is ideal for the high temporal resolution of the MRA, the higher relaxivity was found to be ideal for the vessel assessment in the high resolution MRA. MR-Angioclub East Lansing 2009 71 7.3 Our Strategy for the Surgical Planning with 3T MRA in Detecting Unruptured Cerebral Aneurysms 1) Keiji Igase , Ichiro Matsubara1), Masamori Arai1), Jyunji Goishi1), Hitoshi Miki2), Kazuhiko Sadamoto1) 1) Department of Neurosurgery, Washokai Sadamoto Hoapital 2) Departments of Radiology, Ehime University School of Medicine Purpose: Subarachnoid hemorrhage is the most disastrous and irrevocable disease out of many cerebrovascular diseases except for a few fortunate cases, thus it has been overwhelmingly emphasized to detect unruptured intracranial aneurysms (ANs) with less invasive examination. Therefore, we made our original 3T MRA-centered strategy for the surgical planning on detection of unruptured cerebral ANs and verified the detection capability of them compared with flat-panel detector digital angiography (FD-DA). Methods: In order to screen unruptured intracranial ANs 3T MRI (Signa Excite: GE) was aggressively exploited in our hospital. First of all, out-patients with some cranial problems can undergo 3T-MRI for the screening of cerebrovascular diseases including cerebral ANs, and in the case of patients with ANs suspected on MIP (Minimum Intensity Projection) image of MR Angiography, VR (volume rendering) images are briefly created, with which if ANs are definitely diagnosed, there are two options to proceed treatments depending on their size. For ANs over 5mm in size flat- panel detector digital angiography (FD-DA) is planned for scrutinizing both size and shape of the aneurysms because of its having an indication for the operation, on the other hand for ANs less than 5mm 3D-CT Angiography precedes FD-DA. Objective was all out-patients initially underwent 3T MRI during one year in 2006, where patients with ANs have been followed for 3 years. Results: Out of 3318 of out-patients 286 patients were found to have unruptured ANs, and 3 years after 157 patients were followed, in which 2 patients (1.3%) had the rupture of ANs, 15 patients (5.2%) the operation for unruptured ANs, 19 (12.1%) patients the enlargement of ANs, and 121 (77.1%) no change of the ANs. All of the operated 15 patients underwent both 3T-MRI and FD-DA, revealing that VR images of 3T MRI closely coincided with 3D-images of FD-DA in size and shape of ANs. Conclusion: In detecting unruptured cerebral ANs for surgical treatment 3T-MR Angiography, especially VR images, are sufficiently useful and our strategy with 3T MRIcentered system should be reasonable and beneficial for less invasive examination. 72 MR-Angioclub East Lansing 2009 7.4 Hybrid of Opposite Contrast MR Angiography of the Brain 1 1 2 Faïza Admiraal-Behloul , Evert Blink , Tokunori Kimura , Toshiba Medical Systems Europe, Zoetermeer, the Netherlands, 2MRI Systems Development Department, Toshiba Medical Systems Corp., Otawara-Shi, Tochigi-Ken, Japan 1 Purpose: To compare a new non-contrast enhanced MR angiography technique, to an optimized Time of Flight technique (TOF) with MTC pulse, in the visualization of fast and slow flow brain vessels using comparable scan time on a 1.5T MR system. Methods: The Hybrid of Opposite contrast (HOP) technique is a dualecho 3D gradient-echo sequence where the first echo is used to generate a TOF white blood image and the second echo generates a Flow Sensitive Black Blood image using motion-probing gradients to Figure 1: (a) HOP image (b) TOF image. introduce intra-voxel flow dephasing [1]. In this work, the 2-echo images are combined using a frequency weighted subtraction that enhances slow-flow (small) vessels [1]. Fifteen volunteers (8 men, mean age 46.2 y, range 31 to 77 y) were imaged on a 1,5T System (Vantage ZGV Atlas, Toshiba) using a 13-channel Atlas-head coil with a Parallel Imaging factor of 2. Both TOF and HOP images were obtained using a matrix of 256x256, FOV of 22x22cm, in-plane resolution of 0.86x0.86mm, slice thickness of 1mm and a slab thickness of 7cm. For TOF, TR = 30ms, TE = 8ms and acquisition time = 5min12s. For HOP, TR = 34ms, TE1 = 7.5ms, second TE2 = 24.9ms, and acquisition time = 5min30s. Water excitation technique was used for fat suppression. A slice-selective off-resonance sync pulse was used in TOF or optimal background signal suppression. No MTC pulse was used in HOP. Results: HOP and TOF were visually comparable in the depiction of fast flow vessels in all subjects. The low-flow vessels were better visualized by HOP in 13 subjects (see fig.1 for a representative case) and comparable in 2 subjects. Conclusion: HOP-MRA is a promising technique for the visualization of small collateral vessels which is technically difficult with a standard TOF technique especially at 1.5 T. Reference: 1. T. Kimura et al, Magn Reson Med. 2009 Aug; 62(2):450-8. MR-Angioclub East Lansing 2009 73 7.5 High Resolution Fast Inversion Recovery MRA (FIR-MRA) E. T. Tan, N. G. Campeau, J. Huston III, S. J. Riederer Department of Radiology, Mayo Clinic, Rochester, MN 55905 USA Purpose – Compared to 3D time-of-flight (TOF), the FIR-MRA technique provides superior vessel conspicuity in imaging of the intracranial arteries [1]. However, the intrinsic signal modulation of FIR-MRA causes loss of vessel sharpness and small vessel signal. The FIR angiogram is obtained by a difference between the acquired, un-subtracted bright-blood and black-blood data. Interestingly, the bright-blood signal is larger than the difference signal at k-space periphery, while the dark-blood signal is negligible at k-space center. We hypothesize that these properties of the un-subtracted data may be harnessed to improve vessel depiction of high-resolution FIR-MRA by re-synthesizing the FIR-MRA data. Methods – Two data re-synthesis steps are proposed. First, the use of the difference data for central k-space and the un-subtracted bright-blood 4 cm data for k-space periphery reduces signal modulation and improves sharpness. Second, an additional subtraction of a fraction λ of magnitude dark-blood data with a threshold at zero reduces the level of residual tissue signal. The optimum value for pass band frequency in step one was determined to be 70% of maximum k-space, and that for λ of step two was 10%. This technique was evaluated in A vivo at 3T. Results – Fig. 1 shows improvements in the smoothness and sharpness of large vessels, and in the conspicuity of small vessel branches. The background noise and residual tissue signal are also noticeably reduced. Conclusion – Re-synthesis of FIR-MRA data results in superior vessel depiction, as well as reduction in residual tissue signal and background noise. References – [1] Tan ET et al. ISMRM 2009, #91. B Fig. 1. Axial targeted MIPs (35 mm thick) at the left internal carotid artery before (A) and after (B) resynthesis, with improvements in vessel depiction noted (arrows). 74 MR-Angioclub East Lansing 2009 7.6 3D Dual V E NC P C MR A us ing S piral P rojec tion Imaging 1 N. R. Zwart , and J. G. Pipe1 Keller Center for Imaging Innovation, Barrow Neurological Institute, Phoenix, Arizona, United States 1 P urpos e: The dual-VENC method of unaliasing a low VENC with a high VENC MRI data set allows for a significant improvement in the velocity to noise ratio of PCMRA [1]. The collection of additional data necessitates the use of a rapid imaging technique to maintain feasible scan durations. The proposed method is a 3D k-space sampling trajectory called Spiral Projection Imaging (SPI) [2]. This work demonstrates the effectiveness of SPI while maintaining a short sample window and increased undersampling. The synthesis of high VNR images is explored with attention to problematic areas of dephasing. Methods : S P I: Variable density spiral interleaves are critically sampled up to 1/8 of the k-space radius (Fig. 1a). At the edge of k-space, spirals are separated by a factor of 4 times the critical sample distance. Spiral planes are rotated about kz to fill 3D k-space (Fig. 1b). The scan parameters are: 24cm FOV, 0.8mm3 voxel, 300 dia. matrix, 20 interleaves, 120 planes, 19.5msec TR, and a 6min total scan time for 7 volumes. Dual VE NC : Velocity encoded sets were collected for a 100cm/s and 20cm/s VENC. High velocity gradients at vessel walls cause vessel narrowing in unaliased low VENC images. This is addressed by a weighted combination of the high and low VENC data. R es ults and Dis c us s ion: SPI PCMRA scans were collected with a GE 3T Signa Excite system (Fig. 1c). Blurring near the sinuses becomes problematic with longer sampling windows than what is used. Aliasing due to undersampling causes some smaller vessels to appear inconsistent. Vessel diameter in high flow areas is regained by partial combination of high VENC data with the unaliased low VENC data. C onc lus ion: This method provides short scan times that make the added time required by dual VENC techniques less prohibitive. R eferenc es : [1] Lee, A. MRM, 33:122, 1995; [2] Irarrazabal, P. MRM, 33:656, 1995; MR-Angioclub East Lansing 2009 75 7.7 High Resolution Simultaneous Angiography and Venography (MRAV) with a Single Echo 1,2 Samuel Barnes, MS , E. Mark Haacke, PhD1,2 1. Wayne State University, Detroit 2. Loma Linda University, Loma Linda, CA Purpose: In this work we develop novel acquisition and post processing techniques to improve the quality of single echo MRAV acquisition using susceptibility weighted imaging (SWI). Techniques recently presented have focused on double echo acquisitions which will be harder to implement at very high fields. Methods: All SWI images were acquired at 3T with TR/TE/FA/Resolution/BW = 30ms/20ms/15°/0.5x0.5x0.5mm3/50Hz/pixel or 160Hz/pixel. To visualize veins the data was downsampled to a resolution of 0.5x0.5x2.0mm3 before the SWI processing. Results: Using a higher bandwidth of 160 Hz/pixel as compared with 50 Hz/pixel dramatically reduces flow dephasing, improving larger vessel visibility at the cost of a reduced signal-to-noise ratio (Fig. 1). The high isotropic resolution also helps to reduce dephasing across a voxel improving the MRA. The original isotropic resolution data showed poor venous contrast due to the non-optimal aspect ratio of 1:1 which causes the phase for certain orientations of veins to have the opposite sign (Figure 2). This lost contrast is fully recovered by downsampling the high resolution data to a more optimal aspect ratio of 1:4 (0.5x0.5x2 mm3). The downsampling takes advantage of the high resolution data by reconstructing the slices in an overlapping pattern so the optimal partial voluming of smaller structures is guaranteed thus improving image quality. Figure 1. 50 Hz/pixel (left) shows almost complete loss of the MCA and losses in the middle of smaller vessels while 160Hz/pixel (right) shows minimal flow losses. Figure 2. Isotropic phase image (left) shows veins switching from bright to dark depending on orientation. Downsampled image (right) shows higher SNR and more homogenous veins producing a better venography. Figure 3. Maximum intensity projection (left) over isotropic data showing arteries and minimum intensity projection (right) over SWI processed downsampled image. Conclusion: By collecting data with high isotropic resolution and higher bandwidth, flow related losses from higher order uncompensated effects can be reduced improving the quality of the MRA extracted from the SWI data even at long echo times (20ms). 76 MR-Angioclub East Lansing 2009 7.8 Low Dose 3D Time-Resolved MR Angiography of the Supraaortic Artery: Correlation to High Spatial Resolution 3D Contrast-Enhanced MRA Yoon-Joo Lee, So-Lyung Jung, Kook-Jin Ahn, Bum-soo Kim Department of Radiology, Seoul St.Mary Hospital, The Catholic University of Korea Purpose: To evaluate the effectiveness of low-dose, contrast-enhanced, time-resolved, three dimensional (3D) magnetic resonance (MR) angiography (TR-MRA) in the assessment of supraaortic vessel, and to compare the results with high-resolution contrast enhanced MRA (HR-CEMRA). Materials and Methods: 45 consecutive patients underwent contrast enhanced 3D TR-MRA and high spatial resolution 3D CE-MRA for evaluation of neurovascular disease at 3T. Gadolinium-based contrast medium was administered at a constant dose of 1ml for TR-MRA, and 0.1mmol/kg for HR-CEMRA. Two readers evaluated image quality using a four point scale (from 0=excellent to 3=non-diagnostic), artifacts and findings on both datasets. Results: The overall image quality for low dose TR-MRA was in the diagnostic range (median 0, range 0-3). Two cases showed non-diagnostic image quality due to severe motion in patients with acute ICA occlusion. Readers demonstrated additional hemodynamic information on TR MRA in 3 patients with severe stenoocclusive lesions. For the evaluation of arterial stenosis, TR-MRA well correlated with HR-CEMRA (r=0.668, p<0.001). Of the 675 supraaortic arterial segments evaluated for stenosis or occlusion, TR-MRA agreed with HR-CEMRA in 611 of 675 (90.5%), overestimated in 41 of 675 (6.1%), and underestimated 23 of 675 (3.4%). Conclusion: TR-MRA can be achieved by administration of small contrast dose (1cc, 0.1mmol), and yields rapid and important functional and anatomical information in the evaluation of supraaortic arteries. Due to limited spatial resolution, TR-MRA is has tendency to overestimated the stenosis or occlusion of smaller intracranial arteries. MR-Angioclub East Lansing 2009 77 7.9 Obstruction of IJV by Asymmetry of Lateral Mass of Atlas on Head and Neck CEMRA and Contrast CT 1 Tae-Sub Chung, MD , Hye Mi Gweon, MD1, Sang Hyun Suh, MD1 Department of Diagnostic Radiology, Gangnam Severance Hospital 1 Yonsei University, Seoul 135-720, Republic of Korea Purpose; To evaluate that asymmetry of lateral mass of atlas could be the cause of high level internal jugular vein (IJV) obstruction on head and neck contrast enhanced 3D MR angiography (CE-MRA) and contrast enhanced computed tomography (CE-CT). Materials and Methods; Thirty cases among 1800cases which examined both head and neck CE-MRA and CECT were enrolled during last 5 years. The eleven cases had IJV obstruction and nineteen cases had no IJV obstruction on CE-MRA. We defined obstruction group which had IJV obstruction and control group which had no IJV obstruction on CE-MRA. The following parameter was measured from axial images of CE-CT: 1) the diameter of IJV; 2) the distance between styloid process and ipsilateral lateral mass of atlas; 3) maximum area of lateral mass of atlas. Results; The incidence of IJV obstruction was 28% (504/1800 cases) of all reviewed head and neck CE-MRA. The diameter of IJV and distance between styloid process and lateral mass of atlas at IJV obstruction side in obstruction group were 1.6 ± 1.0mm and 4.1 ± 2.1mm. The diameter of IJV and distance between styloid process and lateral mass of atlas were significantly narrower than those of contralateral normal side in obstruction group (3.5 ± 1.7mm and 3.7 ± 3.2mm, p<0.001). The maximum area of lateral mass of atlas at IJV obstruction side was significantly larger than that of contralateral normal side in obstruction group (24.2 ± 10.5 mm2, p<0.001). Conclusion: The cause of high level IJV obstruction on head and neck CE-MRA was narrowing space between styloid process and lateral mass of atlas resulting from larger area of lateral mass of atlas and induced IJV compression. 78 MR-Angioclub East Lansing 2009 ® 8.1 Evaluation of Gd-DOTA (DOTAREM ) enhanced MRA compared to time-of-flight MRA in the diagnosis of clinically significant non-coronary arterial disease at 1.5 and 3.0 Tesla. 1 Dipan J. Shah , Lim Tae-Hwan2, Steven Wolff3. Methodist DeBakey Heart & Vascular Center, The Methodist Hospital, Houston, TX. 2 Department of Radiology, Research Institute of Radiology, University of Ulsan, College of Medicine, Asan Medical Center, Seoul, Korea. 3 Advanced Cardiovascular Imaging, New York, NY. 1 PURPOSE: To assess the diagnostic accuracy and safety of meglumine gadoterate (GdDOTA)-enhanced MRA over non-enhanced Time-of-Flight (TOF) MRA at 1.5 and 3.0 Tesla for clinically significant non-coronary arterial disease by comparing of each technique with x-ray angiography. MATERIAL AND METHODS: Multicenter, open-label, paired trial in 192 subjects (100 at 1.5 Tesla and 92 at 3.0 Tesla), age >18 years, (140 men, 52 women; mean [±SD] age, 63.7 ± 13.9 years) with suspected non-coronary arterial disease and scheduled to undergo x-ray angiography were included and received an iv bolus of 0.1 mmol/kg Gd-DOTA. Renal insufficiency was present in 15 (7.8%) of subjects. The percent agreement was defined as the number of stenosis grades measured with MRA in agreement with x-ray at segment level / total number of segments evaluated for the patient x 100. Sensitivity and specificity were assessed at the segment level. RESULTS: The arteries imaged were: aorto-iliac (39.6%); renal (18.2%); calf (13.0%); femoral (12.5%); carotid (12.0%); and popliteal (4.7%). There was a statistically significantly greater mean (± SD) percent agreement of MRA to x-ray with Gd-DOTA MRA vs TOF MRA (85.8% ± 19.8% vs 78.3% ± 24.9%, respectively; difference 7.4% ± 22.1% [p<0.0001]). The sensitivity and the specificity of Gd-DOTA MRA were significantly greater than TOF MRA (93.3% vs 85.7% [p=0.0004] and 89.7% vs 84.4%, [p=0.0019] respectively). There were no serious drug-related adverse events and no spontaneously reported cases of nephrogenic systemic fibrosis. CONCLUSION: Gd-DOTA-enhanced MRA provided significantly greater diagnostic accuracy than TOF MRA at 1.5 and 3.0 Tesla in the diagnosis of non-coronary arterial disease. MR-Angioclub East Lansing 2009 79 8.2 Pulmonary MRA in 75 patients with dyspnea ML Schiebler1, SK Nagle1, CJ Francois 1, RF Busse4, ACS Brau4, JH Brittain4, TM Grist1, and SB Reeder 1, 2, 3 1. Department of Radiology, UW- Madison 2. Department of Medicine, UW-Madison 3. Department of Medical Physics, UW-Madison 4. Applied Science Lab, General Electric, Waukesha, WI Purpose: Review of our clinical experience with a new contrast enhanced pulmonary MRA (CE-MRA) sequence with 2D parallel imaging (ARC) 1 for the detection of pulmonary embolism (PE) performed in 75 dyspneic patients. Methods: A total of 75 CE-MRA (accelerated with 2D-ARC) pulmonary MRA exams in patients with dyspnea were reviewed for the presence of emboli (PE), perfusion defects within the lungs, and artifacts limiting diagnostic quality. These artifacts fell into 4 categories: central vessel signal drop-out (CVSDO), central field-of-view (FOV) blurring, bolus timing issues, and residual aliasing. Results: 20 pulmonary emboli were detected in 11 patients. 27 lung perfusion defects were demonstrated. Lobar (n=8), segmental (n=7) and sub-segmental (n=2) PE were identified. Only 5 cases contained artifacts that required follow-up imaging with computed tomography angiography (CTA) to exclude PE. In most cases of CVSDO, use of multiphasic imaging allowed differentiation from PE. Central FOV blurring and residual aliasing were only a problem when the entire AP dimension of the patient was not included in the FOV. Bolus-timing issues were not a significant problem. Conclusion: CE-Pulmonary MRA with 2D-ARC allows near-isotropic high-resolution whole chest coverage in a14s breath hold. This is now being used routinely to evaluate young patients with suspected pulmonary embolism in order to limit radiation dose to this vulnerable population at our institution. Figure 1: Pulmonary embolus in patient with dyspnea. A). Pulmonary infarct (arrow) on pre-contrast 2D ARC, B). Perfusion defect seen on dextro-phase of CE-MRA in left lower lobe (arrow), C). Segmental embolus in LLL lateral segmental artery (arrow) causing the perfusion defect seen in A and B. A B C References: 1. Schiebler et al, ISMRM 2008, Poster # 3928 80 MR-Angioclub East Lansing 2009 8.3 Origin and Frequency of artifacts in Contrast Enhanced Pulmonary MRA in 80 patients with dyspnea ML Schiebler1, SK Nagle1, CJ Francois 1, RF Busse4, ACS Brau4, JH Brittain4, TM Grist1, and SB Reeder 1, 2, 3 1. Department of Radiology, UW- Madison 2. Department of Medicine, UW-Madison 3. Department of Medical Physics, UW-Madison 4. Applied Science Lab, General Electric, Waukesha, WI Purpose: Clinical use of a new contrast enhanced pulmonary MRA sequence with 2D parallel imaging (ARC) 1 for the detection of pulmonary embolism (PE) requires a careful understanding of the normal range of artifacts to help prevent a false positive interpretation. Methods: A total of 80 CE-MRA (accelerated with 2D-ARC) pulmonary MRA exams in patients with dyspnea were reviewed for the presence of significant artifacts (Gibbs ringing, cardiac motion induced blurring, aliasing, respiratory motion) necessitating further imaging with CTA to rule out the presence of PE. Results: Of the 80 exams, only 5 cases had artifacts on pulmonary MRA requiring further imaging with CTA. These artifacts (4 cases) appeared as central hypo-intense foci that resembled PE. The origin of these artifacts may be a Gibb’s ringing phenomenon or from mixing of contrast (from the SVC) with un-opacified blood from the IVC. These occurred most commonly in the right and left lower lobe (LLL) pulmonary arteries. In addition, cardiac motion occasionally obscured the right middle lobe and lingular branches (1 case). In this series, respiratory motion and aliasing were not problematic artifacts. Conclusion: Understanding the artifacts is necessary when interpreting pulmonary MRA. With understanding of these artifacts, highly diagnostic pulmonary MRA can be performed in the vast majority of patients, greatly reducing the use of unnecessary radiation. Figure 1: A. Gibbs ringing artifact simulating a central PE in the left lower lobe pulmonary artery B. Cardiac motion induced blurring of right middle lobe arteries C (arrows). Respiratory motion in a positive case of pulmonary embolism showing perfusion defect (arrowhead) in left upper lobe and a LLL PE (proven at CTA). References: 1. Schiebler et al, ISMRM 2008, Poster # 3928 MR-Angioclub East Lansing 2009 81 8.4 Gadolinium Enhanced Magnetic Resonance Angiography for Pulmonary Embolism: Results of PIOPED III Paul D, Stein, MD Visiting Professor, Department of Medicine, School of Osteopathic Medicine, Michigan State University, East Lansing, MIchigan Purpose PIOPED III is a multi-center prospective study of contrast-enhanced magnetic resonance angiography (MRA) and venography (MRV) accuracy for diagnosis of acute pulmonary embolism (PE). The study rationale was the assumed safety of MRA in the large number of patients who have contraindications to CT angiography (CTA) related to iodine allergy, impaired renal function, or ionizing radiation. Materials and Methods Patients were eligible for the study if they were suspected of having PE. Exclusions included age under 18, inability to complete MRA within 72 hr of the reference test, contraindications to MRA, critical illness, and other standard exclusions. All patients had Wells’ score, MRA and one or more reference imaging tests (CTA, V/Q scan, digital subtraction pulmonary angiography or lower extremity ultrasound). Basis for confirmation of PE included CTA showing PE in main or lobar pulmonary arteries; positive DSA; and either 1] positive CTA (not main or lobar) or positive CTV; or 2] high probability V/Q scan; or 3] positive lower extremity US; in combination with high or intermediate Wells’ score. Basis for exclusion of PE included the inverse of the confirmatory criteria, and patients who did not meet the confirmatory or exclusionary criteria were classified as PE uncertain. Results During the course of the study, a novel disorder (nephrogenic systemic fibrosis) was attributed to use of gadolinium in the presence of impaired renal function, and this impacted recruitment negatively. A total of 818 patients were enrolled. The high prevalence of relative contraindications to CTA noted in the PIOPED II study was confirmed. The image quality, sensitivity and specificity of MRA were assessed. The results will be presented in detail. Conclusions The use of MRA in patients with impaired renal function continues to evolve, while patients with iodine allergy and radiation issues remain candidates for MRA. The role of MRA in diagnosis of PE will be discussed. 82 MR-Angioclub East Lansing 2009 8.5 4D time-resolved MR angiography for non-invasive pulmonary post-embolization AVM patency assessment L Boussel, A Cernicanu, D Gamondes, C Khouatra, V Cottin, D Revel, P Douek. Lyon, France Purpose: In Rendu-Osler disease, post-percutaneous embolization recurrence of pulmonary arteriovenous malformation (AVM) patency is often difficult to assess noninvasively using CT because of its poor temporal resolution. We assess the capability of a post IV Gd 4D time-resolved MR angiography (MRA) sequence to distinguish between patent AVMs and healthy normal vessel by analyzing pulmonary arterial and venous enhancement kinetic. Methods: After IRB approval, 8 patients with 8 documented pulmonary AVMs (3 previously embolized (recurrence) and 5 untreated), prospectively underwent: a thoracic MDCT scanner to localize the AVMs; a pulmonary digital substracted angiography (DSA) to assess AVMs patency and a 4D time-resolved MRA with keyhole and view sharing compression method at 3T (Philips Achieva). MRA was performed after IV injection of 15 ml of Gadovist (Bayer-Schering) at a 2cc/s rate with the following parameters: FOV: 500x350x240 mm, spatial resolution: 1.2x1.2x1.4 mm, keyhole factor: 20%, viewsharing compression: 100%, dynamic scan time (temporal resolution): 1.2 s, total acqusition time: 22.7 s for 6 dynamic images. All images were consensually reviewed by two experienced radiologists. Signal value of cross section AVMs afferent pulmonary arteries and efferent veins were compared to reference arteries and veins located in an healthy pulmonary area at the same distance from the hilum than the studied AVM vessels. The difference in Time to Peak (dTTPav) for each couple artery/vein was calculated. A Mann Whitney test was used to compare dTTPav for AVMs and reference vessels and recurrent and untreated AVMs. Results: A complete thoracic coverage with 6 dynamic time points was obtained for each patient within a single breath hold and with sufficient spatial resolution to analyze distal arteries and veins. dTTPav was significantly smaller in AVM (0.45 +/- 0.9 s) than in reference vessels (3.75 +/- 1.35 s), p=0.0006. No significant difference was found between recurrent and untreated AVMs, p=0.4. Conclusion: 4D time-resolved MR angiography is a promising tool for non-invasive thoracic AVM post-embolization patency assessment. MR-Angioclub East Lansing 2009 83 8.6 Radial Sliding Window MRA in Pulmonary Hypertension 1,2 Timothy J. Carroll , Amir Davarpanah2, James C. Carr2, Michael Cuttica3, 2 4 1 John Sheehan , Sanjiv Shah , and Hyun Jeong 1 2 3 4 Biomedical Engineering, Radiology, Internal Medicine, and Cardiology, Northwestern University, Chicago, IL Purpose: To develop an MRI imaging protocol for the quantification of hemodynamic changes resulting from pulmonary hypertension. Methods: We have developed an approach to CE-MRA which is based on a previously reported radially sampled MR fluoroscopic technique using sliding mask subtraction (1, 2, 3). This allows for better A/V separation which aids in the identification of pulmonary Sliding Window Mask-Mode Subtractions Improves the Depiction of The Distal Branches of the Pulmonary Veins. branches (Figure 1). We tested the hypothesis that bolus transit times are pulmonary patients. (n=17, indicators of hypertension in (a) (b) A series of patients male=8, female=9, <age>=46 ± 9.7) were studied with elevated mean arterial pressure (mPAP> 25 mm Hg) measured in a right heart catheterization study confirmed diagnosis the which of pulmonary arterial hypertension. Figure 1. Mask mode subtraction using (a) a precontrast mask image and (b) a “sliding mask” that is a fixed time lag behind the angiogram. The improved depiction of the pulmonary veins (arrows). ROIs were place to cover the proximal and distal branches of the pulmonary arteries and veins. Arterial and venous transit times were calculated individually as a test of the improved separation of the arterial and venous phases. Results/Discussion: We found that bolus transit times, as measured with our high frame rate pulse sequence, were significantly (p<0.05) increased in pulmonary hypertension patients. Conclusions: CE-MRA of the pulmonary arteries may serve as an adjunct to cardiac catheterization. References: (1) Reiderer et all MRM 1998, (2) Cashen MRM 2007, (3) Jeong MRM 2009. 84 MR-Angioclub East Lansing 2009 8.7 Non-Contrast Enhanced Pulmonary Vein MRI with a Spatially Selective Slab Inversion Preparation Sequence Peng Hu1, Michael L. Chuang1, Kraig V. Kissinger1, Beth Goddu1, Lois A. Goepfert1, Neil M. Rofsky2, Warren J. Manning1, Reza Nezafat1 1 Departments of Medicine and 2Radiology, Beth Israel Deaconess Medical, Boston, MA Purpose: With recent reports of adverse effects of use of contrast agents, non-contrast pulmonary vein (PV) imaging has been of interest [1,2]. We propose a non-contrast enhanced free-breathing ECG-gated 3D thin-slab gradient echo sequence with a sagittal slab-selective inversion for PV angiography. Methods: A sagittal inversion slab was applied prior to data acquisition to suppress cardiac structures adjacent to the left atrium (LA) and PV (Fig. 1) thereby, improving PV/LA conspicuity. The feasibility of the proposed method was demonstrated in 5 healthy subjects. Contrast-to-noise ratio (CNR) between LA and right atrium (RA), ascending aorta and pulmonary artery was measured and compared Fig. 1. Sequence diagram. with conventional non-contrast imaging without inversion. Results: Figure 2 shows improvement of PV and LA conspicuity using our method. Figure 3 shows an example volume view of PV image. Compared to the conventional GRE without inversion, our technique increased the CNR between LA and RA and pulmonary artery by 20 and 4 fold (p<0.01), respectively. Conclusions: The proposed technique enhances the conspicuity of the PVs and LA with minimal loss of SNR. References: 1. Francois et al., Radiology 2009; 250:932-939 2. Krishnam et al., Invest. Radiology, 2009,June 25, ePub. Fig 2: A comparison of images acquired using a conventional GRE (top row) and the proposed technique (bottom row). Fig 3: 3D volume view of PV and LA acquired using the proposed method. MR-Angioclub East Lansing 2009 85 8.8 MRA with the “No Phase Wrap” Grace Choi, Martin R. Prince New York, NY PURPOSE: Aliasing artifact in the phase encoding direction, also known as wrap-around artifact, occurs when patient anatomy extends beyond the field of view (FOV). Tissues outside the FOV wrap around to the opposing edge in the phase-encode direction superimposing unwanted signals on the area of interest which may interfere with diagnosis. Aliasing has become more problematic in abdominal and thoracic MRA utilizing parallel imaging which may wrap the arms onto the middle of the torso superimposing on major vessels. Metallic, e.g. aluminum foil, sleeves have been successful in eliminating wrap-around but heating of the metal and cumbersome application of the foil has prevented this from being a viable clinical option. We have developed sleeves that are convenient to apply and eliminates phase aliasing artifact with negligible heating. MATERIALS AND METHODS: Sleeves were constructed using silver and carbon fabrics of varying mesh geometry and weights and treated with liquid silicone to absorb excess heat. The opening diameter for the shoulder was minimized with an elastic liner to adjust to patient size. Imaging was performed on 1.5T and 3.0T systems (GE HDx 14.0) using an 8-channel body array coil. Coronal and axial single shot fast spin echo (SSFSE), axial steady state free precession and spoiled gradient echo sequences were performed with and without sleeves to identify the optimal fabric. A final design was evaluated on abdominal, thoracic and unilateral upper extremity MR Angiography studies. RESULTS: Phase wrap-around artifact was present on all images scanned without sleeves. Corresponding images using sleeves showed no wrap-around artifact. MRA was possible with smaller FOV without wrap around artifact from the arms on all studies. MRA with parallel imaging was possible with smaller FOV. CONCLUSION: Silver and carbon sleeves effectively eliminate phase wrap-around artifact from arms in patients who cannot tolerate elevating arms overhead for abdominal, thoracic and unilateral upper extremity MRA. 86 MR-Angioclub East Lansing 2009 8.9 Unruptured Intracranial Aneurysms; Detection and Follow-up on 3.0T MRA 1 Hitoshi Miki , K Igase3, I Matsubara3, I Kiriyama2, Kazuhiko Sadamaoto3 1 Department of Radiology, Ehime University School of Medicine, Department of Radiology2 and Neurosurgery3, Sadamoto Hospital, JAPAN Purpose: To evaluate the frequency, size, location and enlargement of unruptured intracranial aneurysms (UIAs) found during screening with 3.0T MR angiography (MRA) at our hospital. Methods: 3D time-of-flight (TOF) MRA was performed for 3,414 cases (1,453 men, 1,961 women, a mean age of 65.7 years) without neurological sign during January to December 2006 at our hospital. All screening MRA was performed with a 3.0T MR system by using an eight-element phased array head coil. Results: The UIAs were found in 286 (8.4%) of the 3,414 cases, 93 men and 193 women, with a mean age of 65.7 years. The locations of the UIAs of single aneurysm cases were as follows: anterior cerebral artery (ACA) including anterior communicating artery (A-com) in 43 lesions, middle cerebral artery (MCA) in 68 lesions, C1 portion of internal carotid artery (ICA) in 44 lesions, C2 portion in 7 lesions, C3-5 portion in 79 lesions, and basilar artery (BA) in two lesions and vertebral artery (VA) in 15 lesions. In 30 cases, multiple UIAs were observed. Aneurysm size varied as follows: 143 lesions (50.5%) were less than 3 mm, 94 (33.2%) were ranged from 3 to 5 mm in size, 34 (12%) from 5 to 7 mm, 11 (3.8%) from 7 to 10 mm, and 4 (1.4%) were greater than 10 mm. The frequency of UIAs on screening 3.0T MRA was higher than that of past screening reports with 1.5T MRA. Especially, the frequency of small aneurysms less than 3mm markedly increased in our study. One hundred forty of 286 cases were followed with serial MRA. Frequency of enlargement was 13.6% (Nineteen cases). Conclusion: 3.0T 3D TOF MRA should be excellent modalities for screening and following up unruptured intracranial aneurysms. MR-Angioclub East Lansing 2009 87 9.1 Comprehensive PC MR Imaging in Congenital Heart Disease Oliver Wieben1,2, Kevin M. Johnson1, Elisabeth Nett1, Ben Landgraf1,2, Scott Reeder1,2, 2 3 3 2 2 Mark Schiebler , Sharda Srinivasan , Carter Ralphe , Darren Lum , and Chris Francois Depts. of Medical Physics1, Radiology2 & Pediatric Cardiology3 - University of Wisconsin Purpose: To further develop and validate highly accelerated radially sampled phase contrast imaging (PC VIPR) for clinical use in congenital heart disease [1]. Methods: 24 consecutive CHD patients (range 9 weeks – 67 years) with a variety of pathology including aortic coarctation (9), bicuspid aortic valve (6), tetralogy of Fallot (5), atrial septal defects (3) among others were scanned at 1.5T and 3 T. Typical scan parameters: imaging volume = 320 x 320 x 180 mm3, readout = 256-320, (1.0-1.25 mm)3 acquired isotropic spatial resolution, VENC = 50-150 cm/s, TR/TE = 8.7/2.8 ms, flip = 10º , retrospective cardiac gating, scan time ~ 10 min (50% respiratory gating efficiency). CEMRA and 2D PC images were used for comparisons when available. Data processing in customized analysis and visualization tools (Matlab & Ensight). Results: PC VIPR data sets were successfully acquired in all patients. All anatomical structures visualized on CE MRA images were identified on the angiograms derived from the PC VIPR images. On a case-by-case basis, additional hemodynamic information was obtained including visualization of flow patterns, flow quantification, and transstenotic pressure gradients. PC VIPR flow measurements were more accurate than the clinical routine targeted 2D PC measurements. Conclusion: The radially undersampled 10 min PC MRI acquisition coupled with the developed post-processing tool has proven to be a versatile tool for imaging in CHD. Coupled with an efficient motion correction it can possibly replace exams in the young currently performed under anesthesia with more patient friendly sedation. e SVC f g RPA L P A posterior view, (b) atrial defect: 1.34 L/min, (c) Fig 1: (a-d) Pulmonary venoblar syndrome – (a) analomous pulmonary venous return “Scimitar Vein”: 0.42 L/min, (d) abnormal systemic artery: flow to the right lung., (e-f) Double inlet left ventricle status post bidirectional Glenn – particle traces and flow waveforms, (g) Aortic coarctation - pressure map. 88 MR-Angioclub East Lansing 2009 9.2 Ultrasound-guided Cardiac Gating for Coronary MRA G. Liu1, R. Walcarius2, X.L. Qi2, A. Dick2, G. A. Wright1 1Medical Biophysics, University of Toronto, Toronto, Ontario, Canada, 2Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada Purpose: Mistakes on the order of tens of milliseconds in the timing of imaging windows can incur significant motion artifacts in coronary MRA. We present a new, ultrasoundbased method for identifying quiescent periods by monitoring the velocity of the interventricular septum (IVS), a correlate of flow into and out of the left ventricle [1]. We demonstrate that, compared with using a cine SSFP pre-scan, our determination of gating parameters produces sharper coronary MRA images. Methods: Right coronary artery (RCA) imaging was performed on four healthy volunteers using a GE Signa 1.5T system. Two sequences were used: (1) T2-prep spiral GRE during 6 to 15s breath-holds with 0.77 x 0.77 x 3.6 mm resolution; and (2) respiratory navigated 3D fat-sat SSFP, with 1.4 x 1.4 x 2.0 mm resolution. ECG gating parameters were determined by, first, ultrasound imaging of the IVS, and second, a SSFP cine of the 4-chamber view [2]. Differences in gating parameters produced different results in RCA visualization. Two experienced observers chose the better images of the two data sets in a blinded, head-to-head comparison. Results: One observer always favoured the MRAs guided by the ultrasound pre-scan. The second observer found image quality to be better for the ultrasound method in 7 out of 8 comparisons, and equal between the two methods in the remaining case. Figures 1 and 2 show some sample comparisons. Figure 1: 2D RCA MRA (74 bpm) ECG gating parameters obtained by (left) an ultrasound pre-scan [onset: 600ms; duration: 53ms], and (right) a cine MR pre-scan [onset: 432ms;duration: 158ms]. Figure 2: 3D RCA MRA (62b pm) ECG gating parameters obtained by (left) an ultrasound pre-scan [onset: 650ms;duration: 158ms], and (right) a cine MR pre-scan [onset:561ms;duration:165ms]. Conclusion: In this study, RCA MRA produced sharper images under the guidance of an ultrasound pre-scan versus the guidance of a MR cine pre-scan. References: [1] Mundigler G et al., JClinBasicCardio 2002 (5). [2] Jahnke et al., Radiology 2006 (239). MR-Angioclub East Lansing 2009 89 9.3 Contrast-Enhanced Whole-Heart Coronary MRA at 3T Using Gradient Echo Interleaved EPI (GRE-EPI) 1 2 1 Himanshu Bhat , Sven Zuehlsdorff , Debiao Li . Northwestern University, Chicago, IL, 2Siemens Medical Solutions, Chicago, IL 1 Purpose: Whole-heart coronary MRA is a promising technique for detecting coronary artery disease; however its major drawback is the long scan time on the order of 10-15 minutes. The goal of this work was to reduce the scan time of whole-heart coronary MRA by using a GRE-EPI [1] sequence at 3T. Methods: 6 echoes (TR = 10.6 ms) were acquired after each RF pulse. To minimize kspace modulations, scan parameters were selected using simulations of the Bloch equation. A dual reference scan phase correction technique was used for accurate echo alignment in the presence of increased off-resonance effects at 3T. The GRE-EPI readout was combined with GRAPPA [2] for a further reduction in scan time. Results: Whole-heart coronary artery images were acquired in 7 volunteers with a spatial resolution of 1.0 x 1.0 x 2.0 mm3 in an average scan time of 2.6 ± 0.6 minutes with an average navigator efficiency of 41.7 ± 9.7%. Fig. 1 shows coronary artery images from 2 volunteers using the contrast-enhanced GRE-EPI (a-c) and GRE (d-f) techniques acquired in separate sessions for comparison purposes. The imaging times with GRE-EPI were 1.6 and 2.6 minutes and those with GRE were 4.2 and 8.7 minutes. Both sequences show similar depiction of the Fig 1. Coronary artery images using a contrast-enhanced GRE-EPI sequence (a-c) and a contrast-enhanced GRE sequence (5d-f), acquired in separate scan sessions. coronary arteries. Conclusion: Compared with current techniques, the proposed GRE-EPI method represents a factor of 3 reduction in scan time and a factor of 2 reduction in contrast dose. References: [1] MRM 30(5):609-16, 1993. [2] MRM 47(6):1202-10, 2002. 90 MR-Angioclub East Lansing 2009 9.4 Feasibility of Whole-Heart Coronary MRA on 3 Tesla Using Ultrashort-TR SSFP VIPR 1 1 1 1 J. Xie , P. Lai , H. Bhat , and D. Li Departments of Radiology and Biomedical Engineering, Northwestern University, Chicago, IL, United States 1 Purpose: The purpose of the work was to evaluate the feasibility of whole-heart coronary MRA on 3.0T using SSFP and verify that ultrashort TR with VIPR allows good coronary MRA image quality with SSFP. Methods: Eight healthy volunteers were studied on a 3.0 Tesla Siemens whole-body scanner during free breathing. An ECG-triggered, navigator-gated SSFP VIPR sequence was used for data acquisition. The imaging parameters were: TR/TE = 3.0 ms/1.5 ms, bandwidth/pixel was 868 Hz, resolution = 1.3*1.3*1.3 mm3.15360~16720 projections, 512 readout points per projection were collected. With the SAR limitation, a flip angle of 50~60 degree was used. Adiabatic T2 preparation scheme was used with a T2-prep time of 40 ms [3]. SPIR (Spectral Presaturation Inversion Recovery) was used to suppress the fat signal. Results: The acquisition time of whole-heart MRA ranged between 9 to 13 min. Both left and right coronary arteries from four volunteers were successfully visualized. Figure 1 is a multiplanar reformatted (MPR) LAD, image RCA, Excellent blood and contrast and observed delineating LCX. between myocardium and no is banding artifacts are present. Average Figure 1. MPR images acquired with ultrashort-TR SSFP VIPR from two volunteers. Note the good delineation of coronary arteries. image quality scores were 3.10 with a SD of 0.41. Conclusion: With non-slab-selective excitation for VIPR, TR could be decreased to 3.0 ms as compared to 4.0 ms usually required for slab-selective SSFP. As a result, no apparent image artifacts were observed in the region of interest and excellent delineation of coronary arteries were obtained in our volunteer studies. References: [1] Bi X, et al. JMRI 2005;22:206-212 [2] Shea SM, et al. JMRI 2002;15:597-602 [3] Nezafat R, et al. MRM 2006;55:858-864 MR-Angioclub East Lansing 2009 91 9.5 Cardiac Imaging: Methods for the Detection of Intramyocardial Fat James W Goldfarb PhD Saint Francis Hospital, Roslyn, NY and Stony Brook University, Stony Brook, NY Purpose: Detection and characterization of myocardial infarction has been shown to be critical for both the right and left ventricles. While the primary technique for detection of fibrosis and necrosis is late gadolinium-enhanced (LGE) imaging, chronic myocardial infarction is often associated with fatty replacement. In this presentation, we will discuss the available methods for detection of intramyocardial fat along with their advantages and disadvantages. Examples from each technique will be presented. Methods and Results: Techniques considered include T1 weighted imaging (native T1 weighting or use of inversion or saturation pulses)(1), bSSFP CINE, opposed phased imaging, water-fat separation (2), chemical fat saturation preparation pulses, and late gadolinium-enhanced water-fat separated imaging(3). Accurate sizing of myocardial fat deposition, may only be achievable using “in-phase” MR images due to the loss of signal at water-fat tissue boundaries in opposed phase images. Opposed phase images may offer increased sensitivity for small fat deposits due to this “artifact” at tissue boundaries. Reliability, ease of use, accuracy, sensitivity, specificity, availability and speed are all important issues for widespread clinical usage. Due to its simplicity, precontrast CT imaging (4) is a quick, high resolution, accurate technique, but does not have the specificity for fat detection as does MR imaging. Though CT techniques are under development, MR imaging currently has better methods for contrast-enhanced infarct detection and the quantitative assessment of ventricular function. Conclusions: MR imaging has many techniques for the detection of intramyocardial fat. Though water-fat separation and CT imaging provide the best images, other techniques may be relevant if detection rather than accurate sizing is needed. 1. 2. 3. 4. 92 Goldfarb J, Arnold S, Roth M, McLaughlin J, Reichek N. Magnetic Resonance Shows Fatty Replacement of Left Ventricular Myocardium after Myocardial Infarction. Circulation 2005; 112:II-470. Reeder SB, Markl M, Yu H, Hellinger JC, Herfkens RJ, Pelc NJ. Cardiac CINE imaging with IDEAL water-fat separation and steady-state free precession. J Magn Reson Imaging 2005; 22:44-52. Goldfarb JW. Fat-water separated delayed hyperenhanced myocardial infarct imaging. Magn Reson Med 2008; 60:503-509. Zafar HM, Litt HI, Torigian DA. CT imaging features and frequency of left ventricular myocardial fat in patients with CT findings of chronic left ventricular myocardial infarction. Clin Radiol 2008; 63:256-262. MR-Angioclub East Lansing 2009 9.6 3D spiral high-resolution late gadolinium enhancement Dana C. Peters, Peng Hu, Reza Nezafat, Warren J. Manning Beth Israel Deaconess Medical Center, Dept of Medicine, Harvard Medical School Purpose: Late gadolinium enhancement (LGE) is peerless among imaging modalities for visualizing scar/fibrosis in the heart. High resolution LGE imaging is an important goal to improve visualization of small regions of scar, such as papillary muscle scar, complex scar related to ventricular tachycardia, and RF ablation scar. Spiral imaging with spectral spatial pulses is attractive for high-resolution LGE because it provides excellent fatsuppression, high SNR efficiency, and good motion ghosting properties. We compared a high resolution 3D spiral LGE sequence with our current 3D Cartesian protocol. Methods: Four subjects were imaged 20-30 minutes after injection of 0.2mmol/kg GdDTPA. Spiral scan parameters include 3D, ecg-gating, and NAV-gating, water-selective RF pulses, 1.3 x 1.3 x 5 mm3 spatial resolution, imaging time = 60 heart-beats (neglecting dead time), 14 interleaves, 14 ms acquisition window, and 38 slice-encodings. The sequence acquires 5 TRs per heart-beat, using a kz-centric acquisition, TRTEq= 23ms/3.6ms/40°. Results: Scans were diagnostic in all subjects. Figure 1 compares the spiral LGE sequence with the Cartesian sequences. The image quality and sharpness is similar to the 3D LGE sequence which took 3x longer. Conclusions: The high resolution 3D LGE spiral approach is promising for visualizing scar in the heart. A B C Figure 1: A) Low resolution 3D LGE (2 x 2 x 5 mm) acquired in 2 minutes. B) 3D Cartesian LGE with 1.2 x 1.2 x 5 mm resolution, in 6 minutes. C) Spiral 3D LGE with 1.3 x 1.3 x 5 mm spatial resolution, acquired in 2 minute. All times assume 50% NAVefficiency. MR-Angioclub East Lansing 2009 93 9.7 Temporal Filtering for Sliding Window Time-resolved Angiography: Beyond Density Compensation Solutions Grabow B, Wu H, Block WF, Samsonov AA University of Wisconsin–Madison, Madison, WI USA Introduction: Methods such as “tornado filtering” [1-2] and KWIC [3] exploit the variable sampling density in k-space trajectories such as radial acquisitions to produce a temporal or parametric series of images from a single acquisition. A gridding reconstruction utilizes a density compensation function (DCF) to limit the aperture of oversampled spatial frequencies to a specific time point or contrast-weighting. The sampling aperture widens with the k-space radius to limit the effects of undersampling artefacts as the sampling density decreases. The filters are poorly compatible with iterative reconstruction because of: 1) suboptimal SNR; 2) mixing of temporal information, especially in smaller objects. We present a method to design such filters for iterative methods that is not based on using a DCF. Methods: The DCF in [1-2] is designed to minimize errors in the spatial point spread function (psf). As the k-space radius increases and data become undersampled, these approaches weight data across the widening temporal footprint [4] evenly, degrading temporal fidelity. We instead use an iterative reconstruction that derives an image estimate fi that fWEf-s i = argmin fi ( designed function fits 1/2 ii Wi the ( acquired ) 2) data s and the encoding matrix E : . The minimization is weighted with a specially that emphasizes the current time frame over all spatial frequencies while minimizing the influence of adjacent temporal data and simultaneously providing optimized image SNR relative to conventional DCF. The implementation provide benefits from parallel imaging by including coil sensitivity information in E . Results and Discussion: The figure to the right shows four simulated contrast-enhancing vessels over a background of linearly enhancing peripheral tissue at a midpoint in scan where the peak vessel enhancement is at the distal (inferior) point of the image. There is no horizontal variation between the vessel enhancement patterns. Notice that that with the traditional filter, the larger vessels show variation vertically across the vessel but the smallest vessel shows no change across the image. The iterative methods show a much closer depiction of the actual enhancement for all widths of blood vessel. The iterative method gives more flexibility in the design of the temporal filter, as the minimization process essentially corrects errors due to variations in sampling density. There is a cost in noise amplification, which ultimately limits increases in temporal fidelity. 1) Barger et al MRM 2002 2) Liu et al IEEE-TMI 2006 3) Song et al, MRM 2000 4) Mostardi and Riederer et al, MRM, 2009. 94 MR-Angioclub East Lansing 2009 10.1 Angiographic and Hemodynamic Assessment of the Hepatic Vasculature in Portal Venous Hypertension using High Resolution PC VIPR Kevin M. Johnson, Oliver Wieben, Chris Francois, Ben Landgraf, Scott B. Reeder Departments of Medical Physics and Radiology, University of Wisconsin, Madison, WI, USA Purpose: Several investigators have proposed the use of non-invasive flow measurement techniques such as PC MR [1,2] for the evaluation of hepatic flow patterns and rates in patients with portal venous hypertension; however, with existing techniques evaluation of the entire hepatic vasculature can be challenging due to long scan times and limited coverage. In this work, we present initial results using an accelerated 3D radial sequence for whole abdomen CINE, 3D PC (PC VIPR). Methods: Both normal volunteers and patients with liver disease were scanned on a 3T MR scanner (MR750, GE Healthcare) using a 32-channel torso array coil. PC VIPR was performed using adaptive bellows respiratory gating with 50% efficiency, 1.25mm isotropic resolution, 32 x 32 x 24cm3 FOV, Venc = 25cm/s, retrospective cardiac gating, for a total scan time of approximately 10minutes. Results: Figure 1, shows example anatomical images derived using PC VIPR. Both arteries and veins are well visualized with high resolution. In Figure 2, vortical flow can be observed in a representative visualization of the portal vein. Figure 1. Example PC VIPR anatomical images of the hepatic artery (left) and portal vein (right). Top images show axial limited MIPs while bottom images show coronal limited MIPs. Conclusion: Whole abdomen PC VIPR allows anatomical and hemodynamic visualization of the entire liver vasculature, the splenic and renal vasculature, making it a very promising for non-contrast enhanced evaluation of the hepatic vasculature in portal venous hypertension. References: 1. Yzet et al. EJR 08 In Press. 2.Stankovic et al. ISMRM 09 #3856 Figure 2. Example visualization on the flow in the portal vein, demonstrating laminar mixing from the splenic vein (blue) and SMV (red) MR-Angioclub East Lansing 2009 95 10.2 Renal MR angiography: multicenter intraindividual comparison of gadobenate dimeglumine and gadofosveset trisodium G Schneider1, M Pasowicz2, J Vymazal3, Z Seidl4, M Aschauer5, 6 7 8 9 M Konopka , D Bilecen , R Iezzi , C Ballarati 1. Homburg University Hospital, Homburg/Saar, Germany; 2. John Paul II Hospital, Krakow, Poland; 3. Na Homolce Hospital, Prague, Czech Republic; 4. Neurologicka Klinika, Prague, Czech Republic; 5. University of Graz, Graz, Austria; 6. NZOZ Slaskie Centrum Diagnostyki Obrazowej, Katowice, Poland; 7. University of Basel, Basel, Switzerland; 8. Università G. D'Annunzio, Chieti, Italy; 9. Hospital Valduce, Como, Italy Purpose: To prospectively compare gadobenate dimeglumine and gadofosveset trisodium for contrast-enhanced MR angiography (CE-MRA) of the renal arteries. Methods: 38 subjects with renal vascular disease underwent a first CE-MRA exam with 0.1 mmol/kg gadobenate dimeglumine followed 3-12 days later by a second exam with 0.03 mmol/kg gadofosveset. For both agents, identical T1w SPGR sequences were used to acquire first-pass (FP) coronal images during breath-hold. In 16/38 patients additional steady-state (SS) sagittal and axial images were acquired with gadofosveset. DSA was performed in 34 patients. Images were evaluated by 3 blinded readers in terms of sensitivity, specificity, accuracy, and positive and negative predictive values (PPV and NPV) for detection of significant (≥51%) renal artery stenosis compared to DSA. Findings were compared using McNemar and Wald tests; assessments of FP diagnostic preference were evaluated using the Wilcoxon Signed Rank test; and reader agreement (kappa []) was determined. A full safety evaluation was performed. Results: Gadobenate dimeglumine was consistently superior to gadofosveset (sensitivity: 76-87% vs 68-76%; specificity: 92-99% vs 91-94%; accuracy: 89-96% vs 86-90%; PPV: 70-94% vs 65-76%; NPV: 94-97% vs 92-94%). Significant superiority for gadobenate dimeglumine was noted by 2 readers for specificity (P≤0.02), accuracy (P≤0.005), and PPV (P≤0.018). SS images provided no additional benefit for gadofosveset. Three-reader agreement was excellent (=0.776-0.855). Readers 1, 2, and 3 preferred gadobenate dimeglumine in 11, 17, and 13 patients and gadofosveset in 5, 4, and 5 patients; no preference was expressed for the remaining subjects. Adverse events were reported for 2/38 [5.3%] with gadofosveset but 0/39 [0%] with gadobenate dimeglumine. Conclusion: Better reader preference and diagnostic performance was obtained with 0.1 mmol/kg gadobenate dimeglumine vs 0.03 mmol/kg of the intravascular blood pool agent gadofosveset for CE-MRA of the renal arteries. 96 MR-Angioclub East Lansing 2009 10.3 FINESS (Flow Inversion-prepared Non-contrast Enhancement in the Steady State): a novel technique for non-contrast renal MRA Manojkumar Saranathan1, Ersin Bayram1, and James Glockner2 GE Healthcare, Rochester MN & 2Dept. of Radiology, Mayo Clinic, Rochester MN 1 Purpose: To evaluate a novel balanced SSFP-Dixon technique for non-contrast MRA of the renal vasculature in a single breath-hold Methods: A 3D dual-echo bipolar readout balanced SSFP pulse sequence with a robust two-point Dixon algorithm1 for fat-water separation was developed. This enabled use of radial fan-beam segmentation in ky-kz. Each fan-beam was acquired after a slab selective 180° pulse that effected venous and background suppression2. The radial fan-beam scheme enabled us to acquire the 3D volume in a single 20-22s breath-hold, eliminating the need for respiratory triggering, which is sub-optimal in some patients. Parameters: 70° flip, TR/TE1/TE2 6.2/1.4/2.8 ms, 256x224 matrix, 35 cm FOV, 2 mm thick, 32-40 slices, TI=900ms. All subjects were imaged on a GE Excite system with a 8-channel torso array coil under an IRB-approved protocol. Results and Discussion: Fig. 1 Right renal artery stenosis depicted using a 22s BH FINESS volume rendering (top) and confirmed on conventional x-ray angiography (bottom). Fig. 2. Right renal artery stenosis depicted using a 23s BH FINESS sequence (top) compared to MR contrast enhanced angiography (bottom). FINESS afforded excellent visualization of stenoses in a short 20-22s breathhold with marked insensitivity to Bo inhomogeneities. Preliminary results are encouraging, and suggest that this technique may have clinical utility for rapid, breath-held non-contrast MRA. Figure 1 Figure 2 Refs: [1] Ma et al. MRM. 52:415-419 (2004) [2] Takei et al. Proc ISMRM, p3420 (2008) MR-Angioclub East Lansing 2009 97 10.4 Magnetic Resonance Angiography of the skin for perforatorbased autologous breast reconstruction I Tiffany Newman, MD , Julie Vasile MD II, Joshua Levine, MD II, David Greenspun, MD II, MSc, Robert J. Allen M.D., A.P.M.C, F.A.C.S II, Minh-Tam Chao B.S.R.T(R) MR I, Martin R. Prince, MD, PhD I I Radiology, Weill Cornell Imaging at New York Presbyterian, New York, NY, United States, II The Center for Microsurgical Breast Reconstruction, New York, NY, United States. Purpose: Autologous breast reconstruction after mastectomy using abdominal and gluteal perforator artery flaps has gained popularity due to preservation of the donor site muscle and function. We evaluate skin MRA accuracy for the preoperative mapping of perforating arteries and flap volume estimation. Methods – Pre-operative MRA on 25 consecutive patients undergoing perforator artery based autologous breast reconstruction was performed at 1.5 Tesla using axial 3D LAVA of the skin overlying abdominal and/or gluteal regions with 20ml gadobenate. Perforator artery size and coordinates relative to umbilicus or top of gluteal crease on 3D MRA were compared to findings at surgery. Reconstructed breast volume estimates from volume rendered MRA images were also compared to weights at harvesting. Results – One hundred twenty-five perforator arteries were found at surgery to be located within 1cm of the coordinates measured on MRA and were surgically verified to be suitable for flap perfusion. Surgery verified the arterial course and caliber through the rectus and gluteal muscles visualized on MRA in 47 of 48 arteries. Volume rendering of 3D MRA accurately predicted breast reconstruction volumes. Figures: Preoperative MRA of the skin of the abdomen (top 2 images) and gluteal (bottom 2 images) regions in different patients undergoing perforator based breast reconstruction Conclusion - This study of 25 patients undergoing breast reconstruction shows that 1.5T MRA safely and accurately identifies precise location measurements and vascular anatomy of abdominal wall and gluteal perforator arteries for guiding autologous flap harvesting. 98 MR-Angioclub East Lansing 2009 10.5 Time-SLIP versus DSA in Patients with Renal Artery Stenosis 1 Isabelle Parienty , Faïza Admiraal-Behloul2 , Francis Jouniaux1 , Guy Rostoker3 2 Centre d’Imagerie du bois de verrière, Antony, France. Toshiba Medical systems, Zoetermeer, the Netherlands. 3Service de Néphrologie, Centre Hospitalier Claude Galien, Paris, France. 1 Purpose: To compare the findings in non-contrast enhanced MRA using the Time Spatial Labeling Inversion Pulse (Time-SLIP) technique [1] to those of Digital Subtraction Angiography (DSA) in patient with significant renal artery stenosis (>60%). Methods: Thirty Patients (12 man, mean age 72 ± 11 y) with renal insufficiency and suspected renal artery stenosis were explored. Time-SLIP images were obtained on a 1.5T MRI system (Vantage, TOSHIBA, Tokyo), using an SSFP sequence with respiratorygating and the following parameters: TR=5.2 ms, TI=1200 to 1800 ms, TE=2.6 ms FA 120, FOV 35x35 cm, Matrix 256X256, Speeder Factor 2 , 35 slices, Fat Sat on, and time= 4.30 min. The image quality was visually assessed by an experienced radiologist and scored as: poor: no contrast in the distal branches but interpretable, moderate: moderate contrast in the distal branches and good: strong contrast from the ostium to the segmental arteries. the degree of stenosis was estimated using measurement tools on a post-processing workstation (GPW, Toshiba Medical Systems). A degree of stenosis of 60% or higher was considered as significant. In all patients with a significant stenosis, a DSA was performed. Figure 1 :. (a) Time-SLIP image (scored as moderate) revealed an ostial stenosis on the right main renal artery estimated at 60%, (see arrow). (b) DSA confirming the findings. Results: The Time SLIP images were scored as good in 24 patients, moderate in 5 patients and poor in 1 patient. We detected 18 significant stenosis in 17 patients. In all 17 patients the DSA confirmed our findings . Figure 1 shows a representative case. Conclusion: Time-SLIP is a reliable technique for renal artery stenosis screening and diagnostic in patients with moderate to severe renal dysfunction. A quantitative comparative study where different degrees of stenosis, from 20% to total occlusion, is ongoing. Reference: 1. D. Utsunomiya et al. Circ J 2008; 72: 1627–1630 . MR-Angioclub East Lansing 2009 99 10.6 Simultaneous Renal Angiography and Perfusion Measurement Using Time-Resolved MRA 1 Katherine L. Wright , Raymond F. Muzic1,2, Nicole Seiberlich2, Yu-Hua Fang1, Stephen R. Yutzy1, Mark A. Griswold1,2, Vikas Gulani2 1 Dept. of Biomedical Engineering, Case Western Reserve University, Cleveland, OH 2 Dept. of Radiology, University Hospitals of Cleveland and Case Western Reserve University, Cleveland, OH Purpose Time resolved angiography with stochastic trajectories (TWIST), a 3D view sharing technique, with GRAPPA is used to achieve sufficient spatial and temporal resolution to simultaneously acquire an angiography exam and perfusion measurement with a single contrast dose. Methods Two normal subjects underwent a TWIST renal MRA exam at 3T (Magnetom Verio, Siemens, Erlangen, Germany) according to local IRB protocol. Imaging parameters: FLASH; 0.05 mmol/kg of Gd-BOPTA (Multihance; Bracco Diagnostics Inc., Princeton, NJ), pA=0.2, pB=0.4 [1], TR/TE/FA=2.5ms/0.95ms/21°, TA=3.7s/volume, Res=1.4x1.4x1.5mm3, FOV=350x284x108mm3, GRAPPA R=3. Contrast dynamics were evaluated for manually segmented tissues using a two-compartment model [2]. Results A representative MIP from a single angiographic frame for Figure 1. MIP from a single time frame. an asymptomatic volunteer is shown in Fig 1. This confirms the feasibility of using this method for renal angiography. Fig 2 depicts time course data and model fits for the same volunteer. The rate constant Ktrans was measured as 4.24 (cortex) and 0.14 (medulla), yielding perfusion measurements of 6.63 and 0.22 ml g-1 min-1, respectively. Conclusions The initial feasibility of this method has been demonstrated. This could have significant clinical benefit, as it could potentially assess perfusion deficits created by renal artery stenosis with a gadolinium dose several times smaller than previously needed forperfusion and angiography measurements [3]. References [1] Song, et al. MRM 2009; 61:12428. Figure 2. Time course data from ROIs and model fits. [2] Tofts, et. al. JMRI 1999; 10:223-32. [3]. Michaely et al. Radiol 2006; 238:586-96. Figure 2. Time course data from ROIs and model fits 100 MR-Angioclub East Lansing 2009 10.7 Assessing Kidney Perfusion using Arterial Spin Labeling and Radial Acquisition for Rapid Characterization of Inflow Dynamics 1 N. Artz , K. Johnson1, Y. Huang1, E. Sadowski2, S. Fain1,2 Medical Physics and 2Radiology, University of Wisconsin, Madison, WI, United States 1 Purpose: Quantifying arterial spin labeling (ASL) perfusion measurements, especially in diseased subjects who may demonstrate a wide range of blood flows, benefits from data at multiple delay times1. This research aims to efficiently acquire data at multiple delay times using a radial approach. Methods: FAIR ASL was performed on a healthy volunteer in a 1.5 T MR scanner. From 0.1 to 2.1 seconds following inversion, a 2D radial balanced SSFP readout acquired unique projections with the following parameters: slice orientation = oblique-sagittal, slice thickness = 8 mm, TR/TE/flip = 5.4/2.7ms/30°, BW = 250.33 kHz, FOV = 34 cm, and matrix = 128 x 128. Control (non-selective inversion) and tag (selective inversion) were alternated until 55 pairs were acquired in 11 minutes. The unique radial lines from all related inversions were combined and partitioned into twenty time frames, each with a temporal window of 55 ms. A cortical ROI was Tag Control Difference used for signal analysis. Results: The perfusion-weighted difference image (at a delay of 1.4s) suffers from poor SNR [Fig 1]. However, the cortical ROI signal vs. time Figure 1 Sagittal control, tag, and difference image at delay time of 1.4 frame curve demonstrates the correct trend for sec. perfusion with the tag showing higher signal than the control [Fig 2]. Further pulse sequence development could reduce the TR, BW and off-resonance effects. HYPR related reconstruction Cortical ROI Signal vs. Time during Blood Inflow techniques may improve SNR and further reduce scan Conclusion: Preliminary results suggest that a radial approach may efficiently acquire perfusion data at Tag Signal (a.u.) time. Time Frames During Inflow Control Null multiple delay times in a clinically feasible scan time. 0.6 1.1 1.6 Delay Time (sec) 1 References: Parkes et al. Magn Reson Med. 2002; 48(1): 27-41. Figure 2 Tag and control inversion recovery signal vs time for a cortical ROI during the blood inflow period. Scan time: 11 minutes MR-Angioclub East Lansing 2009 101 10.8 Imaging Capabilities for Real-time Guidance and Verification of Transcatheter Arterial Chemoembolization (TACE) Procedures WF Block, EK Brodsky, E Bultman, H Wu, A Samsonov, SB Reeder, O Unal University of Wisconsin–Madison, Madison, WI USA Introduction: Transcatheter arterial chemoembolization (TACE) provides targeted delivery of chemotherapy and embolizing agents to liver lesions using X-ray fluoroscopic guidance. The inability of X-ray to adequately visualize the tumor and the tumor’s vascular supply can cause mistreatment, and thus potentially unnecessary damage to healthy liver tissue or incomplete tumor treatment. Our long term aim is to develop an integrated MR real-time imaging system which can visualize liver tumors and adjacent vasculature through dynamic 3D imaging while also verifying the treatment area. Here we describe milestones reached as we move towards system integration. Methods: MR-guided TACE will require catheter tracking, tumor localization, vascular visualization, and treatment validation. Tumor localization has previously been described at the Angio Club. Examining the vascular territory distal to the catheter position is performed via an intra-arterial contrast injection with a 3D stack of stars sequence which can be oriented obliquely in relation to the hepatic vasculature. Verification of the treatment region in TACE is normally performed with a followup CT that is sensitive to the iodine in the ethiodized oil that is preferentially taken up in the treated region. We demonstrate how treatment can be verified immediately by detecting the oil in ethiodized oil in the fat images generated by a single breath-hold MR IDEAL scan. The imaging tasks for this procedure require both real-time, near real-time, and non-real-time capabilities with multiple contrast mechanisms. We are porting catheter tracking (described by Brodsky and Unal at this meeting), vascular imaging, and treatment verification using the virtual scanner capabilities provided by the RtHawk imaging platform (HeartVista, Palo Alto, CA) onto a GE 1.5T HDx system. In the RtHawk paradigm, users exert imaging control through a stand-alone processor which modifies a very simple and flexible imaging sequence. Results and Discussion: Oblique time-resolved CE-MRA of the hepatic vein is shown at 2 s intervals using a stack of stars sequence in Figure a, with time frames advancing left to right. Using an MR IDEAL fat image to validate the treatment region after TACE is shown in Fig. c, where the hyperintense region corresponds to the hyperintense region shown in the CT followup (Fig b). 102 MR-Angioclub East Lansing 2009 11.1 Non-Contrast-Enhanced MR Identification of DVT M Louis Lauzon, Houman Mahallati, Linda Andersen, and Richard Frayne Seaman Family MR Research Centre, University of Calgary, Calgary, AB Purpose: Pulmonary embolism is the 3rd most common cause of death in US hospitals. Common sources of emboli come from pelvic or lower extremity deep vein thrombosis (DVT). We hypothesize that identifying thigh-to-calf DVT is clinically important, so we are investigating high-resolution non-contrast-enhanced (NCE) thrombus MR imaging. Methods: We adapted a 3D direct thrombus imaging sequence1,2 by adding a flowsuppressing bipolar gradient. All images were acquired on a 3.0T scanner (Signa VH/i; GE Healthcare) using a 4-channel torso phased-array coil, coronal orientation, 15º flip angle, 40 cm FOV, 9.2 ms TR, 320×320 in-plane matrix, 90-120 slices for the thigh and calf/knee regions, 2.0 mm slice thickness, velocity suppression of 20 cm/s and above, TE of 2.0/5.4 ms without/with flow suppression, and one signal average, leading to scan times of 4.5 to 6.0 minutes per region. The internal review board approved the study protocol, and each subject gave written informed consent before imaging. Certified body radiologists interpreted the MR thrombus images to determine the presence of clots. Results: The figure below is from a patient with known DVT in the superficial femoral vein. The conspicuity of clot in a given slice is similar without (A) or with (B) flow suppression, but the maximum intensity projection (MIP) image provides significantly better thrombus identification with (D) than without (C) flow suppression. Conclusion: The visualization of lower extremity DVT using high resolution NCE MR imaging is feasible, and shows great clinical promise and potential. A B C D References: 1Moody. Lancet 1997,350:1073. 2Moody. J Thromb Haemost 2003,1:1403. MR-Angioclub East Lansing 2009 103 11.2 Susceptibility mapping as a means to image veins 1,2, 3 2 E. Mark Haacke, PhD and Jin Tang 1. Wayne State University, Detroit, MI, USA 2. McMaster University, Hamilton, Ontario, Canada 3. The MRI Institute for Biomedical Research, Detroit, MI, USA Introduction: The ability to image oxygen saturation is tantamount to being able to follow tissue function in the brain. This is important for monitoring patients with stroke, multiple sclerosis and even tumors. Recently, a new approach to susceptibility weighted imaging (SWI) called susceptibility mapping (1,2) has been proposed. We refer to this new approach as SWIM. This method offers the ability to monitor susceptibility and correlate it with oxygen saturation, the focus of this work. Materials and methods: In order to extract the susceptibility, the phase from a gradient echo sequence is required. In this study, we use the high pass filtered phase image from an SWI scan. This phase image is then Fourier transformed back to k-space, filtered with an inverse regularized filter, and then forward transformed back to the image domain. The resulting image is now a susceptibility map of the veins and the tissue in the brain. High resolution SWI data with isotropic resolution of 0.5mm at 4T were collected. Three echo times were used: 11.6ms, 15ms and 19.2ms. Results: The susceptibility maps for all three echoes give similar results. Although the agreement is not perfect, the susceptibility values appear to be independent of echo time as they should be. They also predict that the oxygen saturation is roughly 0.5ppm in SI units which is consistent with expected in vivo values. Discussion and Conclusion: The ability to map veins throughout the entire brain and to extract venous oxygen saturation is an important adjunct to MRI methodology in the study of diseases that affect the brain’s hemodynamics. We have presented here an approach that promises the ability to provide this information for veins much larger than a voxel. The method is fast and simple and can be applied to any SWI data with phase. References: 1. Deville G, Bernier M, Delrieux J. NMR multiple echoes observed in solid 3He. Physical Review B 1979;19:5666-5688. 2. Salomir R, Senneville BD, Moonen CT. A Fast Calculation Method for Magnetic Field In homogeneity due to an Arbitrary Distribution of Bulk Susceptibility. Concepts in MR Part B (MR Eng) 2003;198:26-34. 104 MR-Angioclub East Lansing 2009 11.3 Modified CAPR MRA: Improved Imaging of the Arterial and Venous Phases P. M. Mostardi, C. R. Haider, N. G. Campeau, S. J. Riederer Department of Radiology, Mayo Clinic, Rochester, MN 55905 USA Purpose – Acquisition parameters for time-resolved MRA are typically constant throughout a scan. We hypothesize that the image quality of the arterial and venous phases of an intracranial MRA can be substantially improved by dynamically changing acquisition parameters, optimizing spatial resolutions and frame rates as dictated by the relevant physiology. Tailoring acquisition parameters to match the specific demands of temporal and spatial resolution of the vascular region of interest is essential in the development of a Comprehensive Neurovascular Exam (CNVE). Methods – The CAPR pulse sequence [1] was modified to allow dynamic change of matrix size, SENSE acceleration, k-space center size, and view sharing factor at a specified time during the acquisition. Volunteer studies were performed in which a viewshared time-resolved view order (2.25 sec/frame, 0.86 x 1.38 x 2.00 mm3) was executed to capture the arterial A phase, and then by seamlessly switching to a high spatial A resolution single-phase view order (25 sec, 0.86 x 0.86 x 1.00 mm3) a venogram was acquired. The reconstruction was modified to automatically account for the change of CAPR acquisition parameters. Results – Fig. 1 shows images from a single timeresolved MRA scan in which several distinct arterial frames were captured as well as a venogram. The arterial B images are optimized for high temporal resolution, whereas the venous phase accentuates spatial resolution and SNR. Conclusion – By dynamically changing view order parameters, the intracranial arterial and venous systems can be imaged with high quality in a single scan. Further enhancements to the CNVE will allow modification of the FOV from a large aorta-based FOV to one limited to the brain. References – [1] Haider CR, MRM 60:3(2008). Fig. 1. Sagittal MIPs of one arterial time frame (A) and the subsequent venogram (B) both taken from the modified acquisition. MR-Angioclub East Lansing 2009 105 11.4 CAMERA: Contrast-enhanced Angiography with Multi-Echo and RAdial k-space 1 Hyun Jeong , Christopher Eddleman2, Saurabh Shah3, Guilherme Dabus4, Timothy J. Carroll1,4 1 Biomed. Engineering, 2Neurosurgery, and 4Radiology, Northwestern University, Chicago, IL 3 Siemens Medical Solutions, Chicago, IL Purpose: A new fast acquisition technique for 4D contrast-enhanced MRA is introduced. It allows a shorter temporal footprint, which samples the contrast bolus more frequently than previous techniques for more accurate dynamic information. Methods: 3D radial “stack-of-stars” k-space is acquired with multiple echoes in partition direction (kz), similar to in a centric segmented EPI (1), which shortens the temporal footprint by up to 60%. Intracranial MRA images of healthy volunteers and AVM patients were acquired on a Siemens 3T Trio, using CAMERA and the previously developed radial sequence. Sliding window reconstruction (2, 3) was used to increase apparent frame rate. AVM patient images were correlated with X-Ray angiography. Results and Discussion: The images acquired with CAMERA (Nechoes=4) resulted in significantly higher CNR values than single-echo acquisition. Flip angle optimization is a known problem with high field CE-MRA, due to SAR. With our multi-echo approach, we are able to increase the TR while reducing temporal footprint. The longer TR used in multi-echo acquisitions allows more flexibility in the choice of flip angles for optimal contrast weighting of the FLASH sequence. Figure 1 shows arterial, nidal, and venous phases of the time-resolved MRA of an AVM, and a corresponding X-Ray angiogram. Figure 1: 4D MRA of an AVM using CAMERA. NRO=192, Nproj=192, Npartition=30 (10% OS), Nechoes=4, FA=45°, TR=6ms, TE=1.5, 2.5, 3.5, 4.5 ms, FOV=220x220x3 mm. Conclusion: A novel fast 4D MRA technique based on radial k-space and multi-echo has been developed. Temporal footprint=9s. Arrows indicate early drainage. References: (1)Beck G et al. MRM, 2001 (2)Riederer et al. MRM, 1988 (3)Cashen et al. MRM, 2007 106 MR-Angioclub East Lansing 2009 11.5 Time-Resolved, Vessel-Selective, Cerebral Angiography Using Arterial Spin Labelling Philip M. Robson1, Weiying Dai1, Ajit Shankaranarayanan2, Neil M. Rofsky1, David C. Alsop1 Beth Israel Deaconess Medical Center, Boston, MA, 2Global Applied Science Laboratory, GE Healthcare, Menlo Park, CA 1 PURPOSE: X-ray Digital Subtraction Angiography (DSA) is the conventional, yet highly invasive method for morphologic and haemodynamic assessment of the cerebral vasculature [1-2]. The purpose of this work is to assess an Arterial Spin Labelling (ASL) MRI method capable of time-resolved inflow visualisation and vessel-selective labelling of feeding vessels, similarly to X-ray DSA, without the use of contrast material. METHODS: Temporally resolved image frames were obtained by varying the duration of labelling before commencing imaging. A modification of pulsed-continuous labelling (pCASL) incorporating additional gradient pulses, allowed labelling to be targeted to the ICA [3]. Labelled signal was acquired whilst in the lumen using a bSSFP read-out. Background suppression was used to provide robust subtraction images. Imaging time was 1.5 min. Quantitative measurements included arterial transit time to vessel segments, residual contralateral signal, and labelling efficiency for vessel-selective labelling. RESULTS: MIP images with temporal resolution of 200 ms (Figure), and high selectivity and efficiency were obtained (5% contralateral signal, 73% relative efficiency). Normal variations of the vasculature were identifiable by ASL-MRA in our cohort of 6 healthy volunteers. CONCLUSION: This non-invasive ASL technique provides high quality angiographic images of the vasculature, able to show haemo-dynamic function; it may be of particular importance for assessing conditions exhibiting altered or differen-tial arterial transit and collateral flow pathways. Left to right: labelling durations of 600, 800, 1000, and 2000 ms REFERENCES: 1) Borisch I et al., AJNR 2003;24(6):1117-1122; 2) Grzyska U et al., Neuroradiology1990;32(4):296-299; 3) Dai W et al., Proc. ISMRM 2008:184. MR-Angioclub East Lansing 2009 107 11.6 Quantifying L umen G eometry from R outine C arotid C E MR A 1 2 3 Payam B. Bijari , Luca Antiga, PhD , Bruce Wasserman, MD , David A. Steinman, PhD1 1 Biomedical Simulation Laboratory, University of Toronto, Toronto, ON, Canada 2 Bioengineering Department, Mario Negri Institute for Pharmacological Research, Ranica, Italy 3 Department of Radiology, Johns Hopkins University, Baltimore, MD, USA Purpose: Recent work has demonstrated a strong relationship between the exposure of the normal carotid bifurcation to disturbed flow and its three-dimensional (3D) geometry [1]. Insofar as 3D contrast-enhanced MRA (CEMRA) is now almost routine, it becomes possible to consider large-scale studies of local risk factors for atherosclerosis. Before this happens, however, it is important to assess the repeatability of these geometric measurements from routine CEMRA. Methods: As part of the ARIC Carotid MRI study’s repeatability protocol, the left or right carotid bifurcations of 60 participants were scanned twice at a mean(±SD) interval of 65±36 days. 3D contrast-enhanced MR angiograms were acquired at 1.5T using bilateral phased-array RF surface coils at the following spatial resolutions: a coronal slab partitioned into 56 2-mm thick slices, zero-padded to 1 mm; and 200-mm2 field-of-view acquired at 256x160, zero-padded to 512x512. Our Vascular Modeling ToolKit (www.vmtk.org) was used to segment the lumen surface rapidly (typically < 5 minutes) and with minimal operator interaction, from which a number of geometric parameters were extracted automatically as described by Lee et al. [1]. Results: Of the 60 pairs, 9 were excluded because one or both carotid bifurcations resisted segmentation by the rapid protocol. Intra-class Parameter ICC correlation coefficients (ICC), tabulated to the right and Angle 0.80 based on analysis of the remaining pairs, revealed Planarity 0.54 excellent repeatability for three of the four geometric Area Ratio 0.93 factors despite the relatively coarse CEMRA spatial Tortuosity 0.89 resolutions, possible repositioning effects, and purposefully rapid segmentation protocol. These results were found to be insensitive to operator’s grading of the segmented lumen surface as excellent (37%), adequate (41%) or poor (22%). Interestingly, area ratio and tortuosity, the geometric parameters that together best predict the burden of disturbed flow [1], were found to have the best repeatability. Conclusion: In most cases 3D level set segmentation and geometric factor extraction can be performed reproducibility, with minimal operator intervention, from non-ideal CEMRA data. This opens up the possibility for cost-effective retrospective or prospective studies on the role of local risk factors in vascular disease. [1] Lee SW, Antiga L, Spence JD, Steinman DA. Geometry of the carotid bifurcation anticipates its exposure to disturbed flow. Stroke 2008 Aug; 39(8): 2341-7. 108 MR-Angioclub East Lansing 2009 11.7 Magnetic source MRI for quantitative brain iron mapping L. de Rochefort, T. Liu, I. Khalidov, J. Liu, B. Kressler, J. Wu, M.R. Prince, Yi Wang Departments of Radiology and Biomedical Engineering, Cornell University, New York Purpose: Iron deposition in the brain can result from cerebral microbleeds (CMB), which may be associated with an increased risk of devastating intracerebral hemorrhage (ICH), especially in patients anticoagulated with medicines such as warfarin. Studies have indicated that the severity of CMB is strongly related to the occurrence of ICH (1). Quantification of iron deposition in CMB may help quantitatively manage the risk of warfarin associated ICH. Currently, dark regions in T2* weighted MRI have been used to identify the presence of iron (2). This hypointensity caused by intravoxel variation of local magnetic fields depends on imaging parameters and source-voxel geometry, may be confused with other signal voids such as those caused by calcium deposits. Method/Results/Conclusions: We propose a novel magnetic source MRI (msMRI) approach to generate quantitative susceptibility maps for quantitative assessment of CMB iron deposits. The average local magnetic field in a voxel is a convolution of a dipole field kernel with the iron magnetization (mass density X B0 X iron susceptibility). So the susceptibility as a tissue material property can be determined by solving the inverse problem from magnetic field to susceptibility source. Unfortunately, straightforward inversion generates no meaningful susceptibility mapping because of severe noise propagation near the zero points of the dipole kernel. We propose to develop a novel robust inversion method by making full use of all information in the T2* gradient echo image data. The phase image (typically neglected in MRI) contains the intravoxel average field information and is used to generate a local magnetic field map. The magnitude image contains intravoxel field variation information and is used to guide the inverse algorithm through a regularization term. We have obtained very encouraging preliminary data indicating that our inverse approach is highly viable for mapping CMB susceptibilities. References: 1. Lee SH, et al. Neurology 2009; 72:171-176. 2. Haacke EM, et al. Magn Reson Imaging 2005; 23:1-25. MR-Angioclub East Lansing 2009 109 11.8 Targeted Glyco-Magnetic Fe3O4 Nanoprobes for Detection and Molecular Imaging of Atherosclerosis Kheireddine El-Boubbou,a,d Medha N. Kamat,a,d David C. Zhu,b Ruiping Huang,c George Abela,c Xuefei Huang*a Department of Chemistry,a Departments of Radiology and Psychology,b Department of Medicine,c Michigan State University, East Lansing MI 48824 Cardiovascular diseases, often associated with atherosclerosis, are the leading cause of morbidity and mortality in the world. Despite the significant progress in cardiology, there remain large unmet needs to early detect atherosclerotic plaques, especially those which are prone to ruptures causing heart attacks and strokes. One of the major causes of such dramatic event is “inflammation” which occurs during early onset of the disease leading to over-expression of adhesion molecules. This initiates an immune response that eventually leads to the formation of the plaque. Such adhesion cell-surface glycoprotein receptors including the cluster of differentiation (CD44) expressed on leukocytes presents a unique opportunity for the detection of the disease in its preliminary state. Our proposed work is based on the surveillance that hyaluronic acid (HA) is upregulated in atherosclerotic lesions and CD44, its principal receptor, is involved in several atherogenic processes. Polyvalent HA is expected to displace native HA from cell surface CD44 to reduce the development of atherosclerosis. Thus, we engineered novel highly dispersed hyaluronic functionalized superparamagnetic iron oxide nanoparticles (HA-DESPIONs) as proficient probes to non-invasively target atherosclerotic plaques using magnetic resonance imaging (MRI). The targeting agents on the external surface of the magnetic nanoparticles will allow the selective labeling of the plaques. Herein, imaging of atherosclerotic plaques in rabbits was successfully examined. Due to the paramagnetic nature of the nanoparticles, their binding with the plaques greatly enhances the contrast from the surrounding tissue, allowing ready plaque detection by MRI. We anticipate that such novel nanoprobes HADESPIONs will not only deepen our fundamental understanding of the molecular and cellular events characterizing unstable atherosclerotic plaques, but also be potentially developed into a highly innovative therapy for atherosclerosis. 110 MR-Angioclub East Lansing 2009 12.1 4D DSA and Fluoroscopy: A New Challenge for MRA? C A. Mistretta, E Oberstar, B Davis, E Brodsky and CM Strother When DSA was first introduced, it was hoped that X-ray angiography could be accomplished with intravenous injections. Ultimately, the overlap occurring in the 2D projections led to the need for repeated injections and image quality was often marginal due to poor SNR. In recent years 3D rotational DSA techniques, based on rotating C-arm gantrys and large area flat panel detectors have been developed. 3D angiographic nontime-resolved volumes are reconstructed following IV or IA injection of Iodine. Image quality is excellent due to the combination of information over the course of a 5-20 second injection and rotation. In recent years accelerated MRA methods have been developed to permit temporal resolution of about 0.75 seconds over isotropic 320 x 320 x320 voxel volumes. These methods have been modified to permit the acquisition of time resolved 4D DSA data sets with matrices up to 512 x 512 x 512 with frame rates of 20 per second using either the intrinsic projection information in the 3D rotational DSA acquisition or using a combination of these data and an auxiliary conventional single projection DSA run using an IV or IA injection. Real-time fluoroscopic catheter information can be embedded within a roadmap formed from any of the 4D DSA time frames and can be viewed from any angle without gantry rotation. Methods Rotational DSA data are acquired using either intra-arterial or intravenous injections of iodine using typical rotation time of 5-20 seconds. Temporal information is embedded in the reconstructed 3D rotational vascular system using multiplicative projection processing (MPP) using time resolved information from either a separate conventional DSA examination or using the intrinsic projections used to form the rotational reconstruction. Whereas conventional reconstruction from projections typically requires a number of projections dictated by the Nyquist theorem, one or two projections suffice to embed the temporal information in the rotational DSA vessels due to the sparsity of angiographic data sets. Shadowing artifacts occur when signals from the time resolved data set are projected through. However the use of biplane acquisition or the use of separated projections from the intrinsic data set are effective in resolving these. In the latter case temporal resolution is reduced in proportion to the required angular separation. Artifacts can also be reduced based on analysis of contrast time curves where shadowing artifacts will appear as anomalies. .4D fluoroscopy is implemented using MPP using a subtracted catheter-only data set that is multiplied into the vascular tree. Results are shown below. Figure 1 shows the AP and Lateral MIP images of the acquired 3D rotational study formed with a 5 second rotation following an intra-arterial contrast injection. The first four frames of the time resolved AP and lateral MIPS through the 4D DSA data set are also shown. The spatial resolution of the images shown has been reduced by a factor of eight relative to what is possible due to memory limitations of the MATLAB software used to reconstruct these images. This study was performed using a single AP conventional DSA acquisition. MR-Angioclub East Lansing 2009 111 Figure 1 Rotational DSA MIPS and MIPS through the 4D DSA time frames. All time frames can be rotated in arbitrary directions. Figure 2 illustrates two orthogonal views of a single fluoroscopic time frame obtained using single plane fluoroscopy. The catheter can be viewed from arbitrary directions without gantry rotation. When biplane fluoroscopy is used the catheter representation is accurate from all views. When a single projection is used the catheter position is constrained to be in the center of the vessel in the orthogonal view since the projection from the acquired view produces a sheet of intensity through the vessel. This may provide no disadvantage since the position of the catheter in a normal fluoroscopic view is also unknown in one direction. In spite of this, the advancement of the catheter tip is well represented. Figure2 Orthogonal views of a single fluoroscopic time frame from single plane fluoroscopic exposure. The catheter is shown in white and can be view from arbitrary directions without gantry rotation. Figure 3 shows a color display of the time to ½ maximum intensity for the vessels contained in one of the 4D DSA time frames. Figure 3 Color display of time to ½ peak. The color scale is in units of 0.25 seconds. 4D DSA is the result of an approximate reconstruction method and typically suppresses parenchymal information, so semi-quantitative impressions of perfusion deficits will have to be inferred based on localized models involving the differences in calculated vascular transit times combined with already available cerebral blood volume maps. 112 MR-Angioclub East Lansing 2009 12.2 MR Angiography of muscular and collateral arteries in peripheral arterial disease: reproducibility of morphological and functional vascular status Bas Versluis1,4, MD; MD; Patty J. Nelemans3, MD, PhD; Joachim E. Wildberger1,4, MD, PhD; Walter H. 1,4 1,4 Backes , PhD; Tim Leiner , MD, PhD Maastricht University Medical Center, Departments of Radiology1 and Epidemiology2 and 3 Cardiovascular Research Institute Maastricht (CARIM), 5Atrium Medical Center Heerlen, 6University Medical Center Utrecht Purpose: Vascular adaptations contribute to the recovery of peripheral arterial disease (PAD). The aim of this study was to determine the reproducibility of the number of arteries in the upper leg as well as arterial flow. Methods: 10 patients with proven PAD (Fontaine stage II) and 10 healthy volunteers were included. All subjects underwent CE-MRA covering the entire muscular volume of the upper legs twice, with a time interval of at least 72 hours. Reproducibility was evaluated in terms of the smallest noticable difference by the repeatability coefficient (RC), the coefficient of variation and the intraclass correlation coefficient between the two scans and the two readers. Results: Interscan RC for the number of vessels was 1.1 for both patients and volunteers, meaning an increase of 1 vessel per measurement plane between two scans would already indicate a significant effect. Interscan RC for flow was 1.5 mL/s for patients and 1.6 mL/s in volunteers. Interreader repeatability coefficient was approximately four times higher, indicating that the same reader is recommended for follow-up studies. Conclusions: Quantification of the morphologic vascular status using artery count and flow measurements proved reproducible in both patients and healthy volunteers. Because of the high reproducibility, CE-MRA might be helpful in quantifying the development of arterial collateral formation in patients with PAD and can help to understand to complexity of this process. MR-Angioclub East Lansing 2009 113 12.3 Multicenter Studies: Lessons Learned from ADNI Matt A. Bernstein, Jeffrey L. Gunter, and Clifford R. Jack Jr Mayo Clinic, Rochester, MN, USA The Alzheimer’s Disease Neuroimaging Initiative (ADNI) [1, 2] is a six-year, 800-subject observational study to assess how well the combined information obtained from MRI, PET, other biological markers, as well as from clinical and neuropsychological assessment can measure the progression of mild cognitive impairment (MCI) and early Alzheimer’s disease (AD). All of the subjects are imaged with 1.5T MRI, and a subset (25%) with 3T MRI. Half of the subjects also receive FDG PET, and a smaller subset of 120 subjects receives PIB PET. A total of approximately 5500 MRI exams are planned over the Execution Phase of the study, which is scheduled to be completed in 2010. All of the image data are readily available via the Internet to any researcher. Details about the ADNI MR imaging protocol and its development process are documented in [2]. A total of 89 scanners with 38 discrete combinations of vendor/field strength/software revision/hardware configuration are supported. Detailed lists of imaging parameters for those configurations are posted and are publicly available at http://www.loni.ucla.edu/ADNI/Research/Cores/ . The wide variety of supported platforms greatly increases the complexity of the management of the study. In this talk, with benefit of hindsight, a few lessons learned about managing the MR portion of a large, multicenter study will be discussed. In particular, the following points will be covered: - Quantitative and automated QC and standardization methods [3,4]. - Advantages of a phased approach, i.e. a “prep phase” or mini-dry run for a large study. - Collaboration with MRI equipment vendors and how it can be mutually beneficial. Preliminary results from ADNI will also be presented, indicating that the sensitivity of MRI methods compares very favorably with PET and neuropsychological testing. The particular experiences reported here relate to an Alzheimer’s disease study, but most of the lessons learned are quite general and apply to large MR angiography multicenter studies as well. 1. Mueller SG, Weiner MW, Thal LJ, et al. Alzheimers Dement. 2005 ;1(1):55-66. 2. Jack CR, Bernstein MA, Fox NC et al., J Magn Reson Imaging 2008 ;27(4):685-91. 3. Gunter JL, Bernstein MA, Borowski BJ et al, Med Phys. 2009 Jun;36(6):2193-205. 4. Mortamet B, Bernstein MA, Jack CR et al, Magn Reson Med. 2009 Jun 12;62(2):365372. 114 MR-Angioclub East Lansing 2009 12.4 Imaging Considerations in Serial Studies of Vascular Disease M Sakamoto, H Kroll, V Rayz, L Boussel, A Martin, and D Saloner Department of Radiology and Biomedical Imaging, VA Medical Center/UCSF Purpose To investigate strategies for extracting quantitative estimates of important descriptors of vascular disease that would permit a determination of compositional and geometric changes in patients with atherosclerotic and/or aneurysmal disease. Methods We selected 18 patients from a large cohort of subjects who underwent two serial MRA studies of the extracranial carotid arteries and 32 subjects with intracranial aneurysms who underwent at least two serial studies generally acquired one-year apart. Quantitative measures of the lumenal volume was measured manually by three independent readers and using a semi-automated algorithm on TOF- MRA, CE-MRA, and T1-black blood MRI. Lumen boundaries were determined either using a visual threshhold or from histogram and/or profile analyses. In order to account for threshholding variation from factors such as coil sensitivity profiles, a reference segment of normal vessel was selected for calibration. Results Excellent inter-reader and intra-reader reproducibility was demonstrated. In carotid disease it was found that appropriate selection of histogram parameters provided essentially equivalent volume measures for all sequences. In aneurysmal disease, it was found that volume values were reproducible to within 3% for diameters larger than 5 mms, to within 10% for diameters greater than 3 mms, and that smaller aneurysms were associated with substantial uncertainty. Conclusions The combination of MRI and MRA provides information on the temporal evolution of both the lumenal geometry and disease of the vessel wall such as atheroma or juxtalumenal thrombus. However, the absence of normalized values for signal intensities compromises the ability to unambiguously define vessel wall composition and the precise location of vessel edges. With careful attention to signal intensity characteristics and using appropriate reference standards it was found that reliable estimates of serial changes could be extracted from MRI/A. These approaches have significant advantages in longitudinal studies. One specific example is the ability to use MRA studies to quantitate the vascular lumen thereby reducing the imperative to obtain high quality black-blood lumenal definition in studies of atherosclerosis. MR-Angioclub East Lansing 2009 115 12.5 Towards Abdominal MRA at 7 Tesla Ladd ME, Umutlu L, Maderwald S, Kinner S, Orzada S, Antoch G, Kraff O, Ladd SC, Brote I, Bitz AK, Schaefer L, Quick HH, Lauenstein TC 1 Erwin L. Hahn Institute for MRI, University Duisburg-Essen, Essen, Germany Department of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital, Essen, Germany 2 Purpose: Imaging in anatomic regions with large cross-section is challenging at 7T because of RF wavelength effects in the tissue. Thus far, 7T imaging has therefore been primarily limited to the head or extremities. Only few reports of 7T for cardiac or abdominal imaging are available [1-3]. The purpose of this study was to perform an exploratory study of the potential of 7T for performing MRI and MRA in the abdomen. Methods: All examinations were performed on a 7T whole-body MRI system (Magnetom 7T, Siemens) in a total of 20 subjects. A custom-built flexible 8-channel RF transmit/receive body coil consisting of stripline elements and suitable for static RF shimming was used [4]. Non-contrast-enhanced MRI was performed including T1w 3D FLASH, T1w fat-saturated 2D FLASH, 2D T1w in- and opposed-phase FLASH, 2D TOF, quasi T2w 2D TrueFISP, and T2w TSE. Results: T1w imaging at 7T in general revealed excellent conspicuity of small anatomical structures and the hepatic and renal vasculature (Fig. 1). The arterial system was bright and the venous system primarily dark regardless of slice orientation. TrueFISP provided unexpectedly good image quality (Fig. 2); however, both it and TSE remained challenging due to SAR restrictions. Conclusions: These results are promising for the future of performing abdominal MRA at 7 Tesla. The inherently high signal of the arterial system in T1w imaging shows the potential for MRA without contrast agent. Subsequent studies in healthy volunteers and patients will further assess these sequences and the 7T appearance of pathologies. Fig. 1: Non-enhanced T1w imaging. Left and middle: FLASH 2D of the upper abdomen. Note the inherently high vasculature signal. TR/TE = 130/3.6 ms, flip 70°, BW 2 405 Hz/pixel, FOV 400x400 mm , matrix 512x512, 13 slices, thickness 2 mm, Grappa R=2, TA 31 s. Right: TOF MIP in a further subject. Fig. 2: TrueFISP provides good overview of the liver vasculature. TR/TE = 3.5/1.5 ms, flip 50°, BW 975 Hz/pixel, 2 FOV 400x400 mm , matrix 256x320, 21 slices, thickness 4 mm, TA 18 s. References: [1] Vaughan JT, et al. MRM. 2009;61:244-8. [3] Maderwald S,et al. ISMRM 2009, p. 821. [2] Snyder CJ, et al. MRM. 2009;61:517-2 [4] Bitz AK, et al. ISMRM 2009, p. 4767. 116 MR-Angioclub East Lansing 2009 1,2 12.6 7 Tesla Cardiac MRI in Humans Harald H. Quick , S.Maderwald1, S.Orzada1, A.K.Bitz1, I.Brote1, O.Kraff1, L.C.Schaefer1, M.E.Ladd1,2 1 Erwin L. Hahn Institute for MRI, UNESCO World Cultural Heritage Zollverein, Essen, Germany 2 University Hospital Essen, Department of Diagnostic and Interventional Radiology, Essen, Germany Purpose: Human cardiac MRI at 7 Tesla is a potentially challenging endeavor due to inhomogeneous RF signal transmission [1] caused by the heart’s position deep within the body and due to the reduced Larmor wavelength at 7 T of approximately 12 cm, which is shorter than the dimensions of the human body and may thus lead to destructive B1 interference (signal voids). Additionally, the specific absorption rate (SAR) at this field strength often constrains the choice of imaging sequence parameters [2]. The purpose of this study was to transfer initial experiences in animal cardiac MRI at 7T [3, 4] to human in vivo cardiac imaging using a custom-built flexible 2x4-channel RF transmit/receive body coil. Methods: All examinations were performed on a 7T whole-body MRI system (Magnetom 7T, Siemens, Erlangen). A custom-built flexible 2x4-channel RF transmit/receive body coil for 7T human imaging was used for RF signal transmission and reception. Seven healthy volunteers (4 male, 3 female) were placed head-first supine with the chest at the isocenter of the magnet and within the sensitive region of the coil. The imaging protocol encompassed cardiac function along standard views using peripheral pulse-triggered Cine FLASH sequences with 20 phases per RR-interval. The image quality was visually assessed for signal homogeneity and myocardium-to-blood contrast. Results: All seven subjects could be successfully examined. The coil, driven in CP mode, qualitatively provided relatively homogeneous B1 signal over the sensitive volume. Some regions in the images, however, showed destructive interference with associated signal voids. The Cine FLASH sequence provided good imaging quality and signal homogeneity over almost the entire heart and with good myocardium-to-blood contrast; the achieved spatial resolution was 1.4 x 1.4 x 4 mm3. For perfectly timed and triggered cardiac images, however, ECG triggering seems mandatory. Peripheral pulse gating, as used in this study, in part was associated with imprecise triggering leading to mild motion artifacts. Fig. 1: Cine FLASH images of the human heart in vivo at 7T. a) short axis, b) 2-chamber, c) 4 chamber, d) LVOT, and e) LVOT 2nd. 1 out of 25 cine phases is shown for each orientation. Conclusion: These initial results can be considered as a first step towards human in vivo cardiac imaging at 7 Tesla high-field MRI. Subsequent studies in patients will further assess these sequences in high-field cardiac imaging and additional cardiac protocols including late enhancement. References: [1] DelaBarre et al.; ISMRM 2007, p. 3867; [2] Maderwald et al.; ISMRM 2008, p. 2716.[3] Quick et al.; MRA-Club 2008, Graz, p. 84.;[4] Quick et al.; ISMRM 2008, p. 1023. MR-Angioclub East Lansing 2009 117 12.7 The Role of Cholesterol Crystals in Acute Cardiovascular Events: Identifying the Cause for Gender Differences in Clinical Presentation George S. Abela MD, Ameeth Vedre MD, Fadi Shamoun MD, Majid Moughal MD, Department of Medicine, Division of Cardiology, Michigan State University, East Lansing, MI Purpose: Plaque rupture has been seen more frequently in men and erosion in women. To explain this we evaluated the physical factors related to cholesterol crystallization. Methods: The amount of cholesterol, temperature, hydration and pH were varied and rate and amount of volume expansion measured in vitro. Cholesterol powder (1,2,3 g) was dissolved in corn oil at varying temperatures (22-44°C) and allowed to crystallize. Effect of pH and hydration of cholesterol were also evaluated. Fibrous membranes (rabbit pericardium) with similar thickness and composition to the plaque cap were placed in the path of growing crystals to assess potential for damage. Tissue preparation for scanning electron microscopy (SEM) was performed without ethanol to avoid dissolving the cholesterol crystals. Results: The volume and rate of cholesterol expansion was directly related to the amount of cholesterol present (r=0.98; p<0.01, r=0.99; p<0.01 respectively). Low temperature, hydration and higher pH all significantly increased volume expansion. By gross examination, the expanding crystals tore the fibrous membrane while in others it just perforated the membrane. By SEM cholesterol crystals were seen perforating the fibrous membranes. Conclusions: Greater amounts of cholesterol result in more crystallization and volume displacement that can tear fibrous membranes. Thus, the larger necrotic cores seen in men that probably have higher cholesterol content would result in more rupture. This may help explain in part the gender difference in clinical symptoms of more ruptures in men compared to more erosion in women. 118 MR-Angioclub East Lansing 2009 P.1 CE-MRA with tailored 3D random sampling patterns and nonlinear parallel imaging reconstruction 1 2 3 F. Knoll , C. Clason , F. Ebner , M. Aschauer3, R. Stollberger1 Institute of Medical Engineering, TU Graz, Austria, 2Institute for Mathematics and Scientific Computing, University of Graz, Austria, 3Department of Radiology, Medical University Graz 1 Introduction: Variable density 3D random sampling trajectories which were introduced in the context of compressed sensing [1], have great potential for subsampled MR angiography techniques. The goal of this work was to present a parameter-free method to construct tailored variable density sampling patterns, which can be used together with a nonlinear parallel imaging method [2, 3] to allow the use of very high acceleration factors. Methods: A CE-MRA dataset (3D Gradient Echo Sequence, TR/TE=3.74/1.48ms, FA=30°, matrix: 448x352x40, resolution: 0.55x0.55x0.70mm3) of the carotid arteries was acquired on a clinical 3T system and subsampled retrospectively, to simulate an accelerated acquisition. Parallel imaging with 8 receiver coils, comparable to the spatial positions of the individual elements of the receiver head coil that was used in our experiments, was simulated by use of Biot Savart’s law. The proposed method to generate the variable density 3D random sampling pattern is to use the scan of the same anatomic region of a different patient or a healthy volunteer as a template. The power density spectrum of this template can be used as a reference to generate a probability density function which is then used to construct the sampling pattern. Patterns that are generated in this way can be pre-computed for different types of scans or anatomical regions. Image reconstruction was performed with an iteratively regularized Gauss-Newton method (IRGN) [2]. Results and Discussion: Our results show that excellent image quality can be achieved even for very high acceleration factors like R=30 (Fig.1.) without any application of temporal view sharing. Only a slight decrease of SNR and a minimally reduced contrast for the smallest vessels result for an acceleration factor of 30. The mean RMS difference to the original fully sampled data set over all 40 slices was 0.055 with a standard deviation of 0.036. The proposed approach to generate the variable density sampling pattern ensures the sampled ratio of low to high frequency sample points is reasonable for angiography scans. One major advantage is that the method is completely free of any user defined parameters. Our experiments showed that the method is robust regarding the choice of the reference image, as the only information that has to be obtained is an estimate of the ratio of high to low frequency components. The exact anatomical details are not important in this context. Fig. 1: CE-MRA Dataset of the carotid arteries: (a) Fully sampled data set (b) IRGN reconstruction, R=30 References: [1] Lustig et al., MRM 58: 11821195 (2007), [2] Uecker et al., MRM 60: 674-682, [3] Knoll et al., ISMRM 2009: 2721 MR-Angioclub East Lansing 2009 119 P.2 Handling Motion in Sparse MRI with Whiskers Jason Mendes, Dennis L. Parker, University of Utah, UCAIR, Salt Lake City, Utah PURPOSE: For dynamic images, pixel intensities that vary smoothly or periodically can be sparsely represented in a Wavelet-Fourier space and recovered from undersampled kspace data using Compressed Sensing (1,2). Patient motion that is not periodic or smooth can therefore be a problem in the reconstruction. We investigate the application of Compressed Sensing to a novel segmented data sampling technique called Whiskers (for the whiskers looking pattern in k-space and lack of a clever acronym) to reduce the effects of patient motion. METHODS: In general, the minimum number of k-space samples required to produce good results in sparse reconstruction is approximately four times the number of sparse coefficients (3). It is therefore beneficial to detect and correct as much patient motion as possible to maximize temporal sparsity and thus reduce the total number of k-space samples required. This is accomplished using a hybrid Radial-Cartesian sampling technique called Whiskers. The Whiskers k-space trajectories of a single segment are shown in Figure 1. Readout and phase encoding directions are alternated each TR. Figure 1: Whiskers segment (a) combined with segment (b) yields a radial like coverage while sampling the central part of k-space rectilinearly. (a) (c) (b) (c) RESULTS AND CONCLUSION: The new Whiskers design has been implemented with a Turbo Spin Echo sequence and shown to provide good detection of both translational and rotational motion. Aliasing artifacts due to Whiskers undersampling are similar to that of Radial undersampling, but more incoherent, making Whiskers well suited to be used in conjunction with Compressed Sensing. Additionally, the most distinct artifact of Compressed Sensing is the loss of low-contrast features in the image (1). Because Whiskers fully samples the central portion of k-space it should be able to recover much of that loss. The simulated results look promising and the authors intend next to apply the technique to vascular imaging. ACKNOWLEDGMENTS: 1. Lustig M, et. al. Magn Reson Med 2007;58:1182–1195. 2. Lustig M, et. al. ISMRM, Seattle, 2006. p 2420. 3. Tsaig Y, et. al. Signal Process 2006;86:533–548 120 MR-Angioclub East Lansing 2009 P.3 Improved Carotid imaging with HASTE using a reduced FOV and increased gradient performance Jordan P. Hulet5, Seong-Eun Kim1,3, Gerald S. Treiman2,4, Dennis L. Parker1,3,5 Utah Center for Advanced Imaging Research, 2VA Salt Lake City Health Care System, 3Radiology, 4 Surgery, 5Biomedical Informatics, University of Utah 1 Purpose MR imaging of the carotid artery often suffers from motion artifacts. Single shot sequences, such as HASTE, can reduce the effect of motion artifacts but often produce blurry images due a long echo train. We have explored the use of both a reduced FOV technique (shown below) and increased gradient performance in order to shorten the echo train length, decrease blurring, and improve image quality. Methods The standard Siemens HASTE sequence was modified to image a reduced field of view (rFOV) in the phase-encoding direction while preventing aliasing artifacts by moving the gradient waveforms for the 180° refocusing pulses from the slice select to the phase encode axis and adjusting gradient amplitude to accommodate the reduced field of view. A volunteer was scanned using both regular HASTE and rFOV-HASTE. Results Images obtained using both sequences are shown in Figure 1. rFOV-HASTE resulted in images with significantly less blurring which more clearly visualize the carotid arteries. Figure 1: Carotid scans using regular (a) HASTE and (b) rFOV-HASTE. The HASTE images were cropped to match the reduced FOV (25%) of the rFOV-HASTE images. Discussion Using rFOV with HASTE shortens the echo train resulting in improved carotid images. Additionally, a custom gradient insert designed to increase gradient performance is currently under construction. The rFOV-HASTE sequence has been modified to utilize this increased gradient performance and further shorten the echo train. MR-Angioclub East Lansing 2009 121 P.4 Towards Continuously Moving Table NCE Peripheral MRA Randall B. Stafford1,5, Mohammad Sabati4, Houman Mahallati2,5, Richard Frayne1-3,5 1 2 3 Departments of Physics and Astronomy, Radiology, Clinical Neurosciences, University of Calgary, Calgary, AB, Canada; 4Department of Radiology, University of Miami, Miami, FL, USA; 5Seaman Family MR Research Centre, Foothills Medical Centre, Calgary, AB, Canada Purpose: Peripheral arterial disease (PAD) requires large field-of-view (LFOV) angiography. PAD patients with renal artery stenosis may be at risk of NSF [1,2]. The purpose of this project is to combine the balanced steady-state free precession (bSSFP) Dixon method [3,4] with the continuously moving table (CMT) technique [5,6] for LFOV NCE peripheral MRA. Here, we assess the feasibility through computer simulation. Methods: A peripheral anthropomorphic phantom consisting of fat, marrow, arterial and venous blood, and muscle was constructed in MATLAB (The MathWorks, Inc.). Arterial segments were modified with varying levels of stenosis. LFOV bSSFP Dixon method [3,4] image data sets were generated using bSSFP magnetization evolution with a table velocity of 0.98 cm/s. Stationary images were generated for difference comparison. Maximum intensity Fig. 1: Simulated bSSFP Dixon method MIP images of anthropomorphic phantom (left) stationary data set, (centre) CMT data set, and (right) difference. Arrows indicate stenoses. projection (MIP) images were produced for all data sets. Results: The water-only MIP images [3,4] are shown in Fig. 1. The simulated table velocity (0.98 cm/s) allowed the tissue to achieve steady-state magnetization, and allowed for 100% hybrid-space coverage [5]. The blurring artifacts in the CMT images are due to partial-volume effects associated with the coarse phantom resolution and the table motion. This blurring effect would be reduced in a real CMT acquisition [5,6]. Conclusion: The simulations presented here have successfully demonstrated the compatibility of the bSSFP Dixon method with the CMT technique. The next step in this project is to collect LFOV images in vivo, and assess vessel conspicuity. References: [1] Albers, Am J Kidney Dis 1994;24:636, [2] Thomsen, Eur Radiol 2006;16:2619, [3] Huang, MRM 2004;51:243, [4] Stafford, MRM 2008;59:430, [5] Kruger, MRM 2002;47:224, [6] Sabati, PMB 2003;48:2739. 122 MR-Angioclub East Lansing 2009 P.5 R2* Calibration Phantoms for Cardiovascular Studies Matthew T. Latourette and James E. Siebert, Michigan State University, East Lansing Purpose: R2* quantitation has been applied to assess iron overload and, recently, to characterize hemorrhage in atherosclerotic plaque.1-3 Calibration reference phantoms exhibiting long-term chemical stability and temperature-insensitive R2*/T2* signal, needed for longitudinal studies, have not been adequately addressed in the literature. This study aims to improve R2* quantitation reproducibility via development of a calibration reference. The availability of a stable reference standard is important for R2*-based research studies for monitoring data quality, decreasing the variance of pooled intra-site study data, and detecting and correcting bias in multi-site studies. Methods: A gel mixture matching the T1 and T2 of pediatric brain was doped with SPIO nanoparticles (Feridex) to produce phantoms with varying R2*. NiCl2 was used to modify T1 because its relaxation behavior is largely independent of temperature, unlike other popular doping agents.4,5 Agarose was used to modify T2 and promote gel formation. Carrageenan aided formation of strong, tissue-like gels.6 A methylisothiazolinone-based (MIT) preservative provided long-term antimicrobial protection. Heat and stirring was applied to the mixture of NiCl2, carrageenan, agarose, and MIT to dissolve it in water. The solution was poured into polysulfone bottles, chilled, and sealed. Imaging was performed at 0.7T to verify the T1 and T2 of these phantoms. T1 was computed by fitting the mean ROI signal intensity for each of 5 axial SPGR series to S(TR)=S0(1-e-TR/T1). Mean signal intensity in 6 axial spin echo series ROIs was fit to S(TE)=S0e-TE/T2 to compute T2. R2* calibration phantoms were prepared with the same underlying composition, varying only the SPIO concentration. Results and Discussion: Average T1 and T2 for the SPIO-free phantoms at 0.7T was 838 ms and 89.5 ms respectively. R2* as a function of SPIO concentration of the doped phantoms was 38.5±1.3 s-1 at 63.6 µM, 44.4±1.7 s-1 at 72.7 µM, 55.5±1.9 s-1 at 90.9 µM, and 68.6±1.9 s-1 at 127 µM. Follow-up imaging results describing measurement reproducibility, R2* measurements for other field strengths, and an evaluation of the thermal stability of phantom R2* will be presented. References: 1Siebert JE et al, MR Angio Club, 2006. 2Zhu DC et al, ISMRM, 16:2841, 2008. 3 Zhu DC et al, ISMRM, 17:606, 2009. 4Kraft KA et al, MRM, 5:555-562, 1987. 5 Waiter GD, 6 Foster MA, Magn Reson Imaging, 15:929-938, 1997. Yoshimura K et al, MRM, 50:1011-1017, 2003. MR-Angioclub East Lansing 2009 123 P.6 Pictorial Review of Supra-Aortic Artery Pathologies as Visualised with MRA using Blood Pool Contrast Agent Bethapudi S & Roditi G Glasgow Royal Infirmary, Alexandra Parade, Glasgow G31 2ER Purpose: Review the appearances of carotid, vertebral & subclavian artery disease as demonstrated by Contrast-Enhanced Magnetic Resonance Angiography (CE-MRA) with blood pool contrast agent. To understand the utility of high resolution 3D steady state phase imaging in adding value to conventional first pass CE-MRA. Background: The blood pool contrast agent Gadofosveset trisodium allows first pass CEMRA at high spatial resolution with low gadolinium dose due to its high specific relaxivity in human serum. The prolonged blood pool residency further allows very high spatial resolution imaging during the steady state phase with acquisition of isotropic voxels that allow reformat in any plane without loss of spatial resolution along with depiction of adjacent structures not seen on usual CE-MRA. Procedure Details: Our experience of MRA in patients imaged with blood pool contrast agent for suspected carotid, vertebral or subclavian disease is presented. Patients were imaged in usual first pass and subsequently in steady state phase with high spatial resolution. Results are presented as a pictorial review of the spectrum of pathologies encountered including atherosclerotic steno-occlusive disease, subclavian steal syndrome, vertebral dissection, carotid body tumour, surgical endarterectomy sites, large vessel vasculitis, thoracic outlet syndromes and depiction of extra-anatomical bypass grafts. Particular emphasis is placed on the utility of high spatial resolution steady state imaging to extend coverage, delineate the vascular wall and evaluate associated structures. Conclusion: Blood pool contrast-enhanced MRA of the supra-aortic vasculature is effective in allowing comprehensive evaluation of various arterial pathologies, particularly in demonstrating extraluminal pathology. 124 MR-Angioclub East Lansing 2009 P.7 Robust Clinical Application of Time-Resolved MRA K. A. Blackham, G.S Sandhu, R.C. Gilkeson, M.A. Griswold, V. Gulani1 Department of Radiology, University Hospitals of Cleveland, Cleveland, Ohio, United States Purpose: To present a review of the clinical applications of time resolved MR angiography (trMRA) throughout the body. Methods: A series of patients were imaged with dynamic contrast enhanced TWIST-MRA (Siemens Verio, 3.0 T, TR/TE=minimal ; representative 3.28ms/1.51 ms; matrix 256320 x 176-256 x 64-80, GRAPPA 2-3, flip angle 21-25˚, TA 1.54 s/ frame, # frames varied according to application). Results: The TWIST sequence provides a robust clinical tool for trMRA that can be employed to solve clinical problems noninvasively. In children, applications included assessment of arterial inflow and venous outflow in arteriovenous malformations (AVMs), mapping of thoracic and abdominal anomalous vessels, and assessment for renal artery stenosis. The patients were free-breathing, and resolution as high as Fig.1: Sub-volume MIP of a contrast enhanced trMRA in a child showing a thoracic kidney in the right hemithorax and an aneurysm in the renal artery. 0.9x0.7x0.9 cm3 was routinely achieved with a single dose of Magnevist contrast. Applications in adults included AVMs in the brain and throughout the body with mapping of feeder/draining vessels, chest pulmonary angiography and venography, assessment of renal artery stenosis in patients unable to provide breathholds, assessment of AV fistulas, dynamic imaging of popliteal artery entrapment syndrome, pelvic congestion syndrome, assessment of arterial anatomy and disease when contrast dynamics in two legs were drastically different, and assessment of perforator anatomy for fibular free-flap transfer operations, often with half the standard dose of contrast (by weight). Near isotropic datasets allowed multiplanar reconstruction of images at multiple timeframes, and dynamic MIP images allowed visualization resembling traditional angiograms. Conclusion: TWIST-MRA is a rapid, robust technique that is non-invasive and can be used routinely and for problem solving in a variety of settings throughout the body. References: 1. Petkova M, et al. J Magn Reson Imaging 2009; 29:7-122. 2. Sandhu G et al. Proc ISMRM 2009; 589 MR-Angioclub East Lansing 2009 125 P.8 Reproducibility of Aortic Pulse Wave Velocity Measurements Obtained with Phase Contrast Magnetic Resonance (PCMR) and Applanation Tonometry Jonathan Suever, BS,† David Huneycutt, MD,* Enrique Rojas-Campos, MD,* Francesca Cardarelli, MD,* Sam Fielden, BS,* Arthur Stillman, MD, PhD,* Paolo Raggi, MD,* John N. Oshinski, PhD,*† †Emory/Georgia Tech, Department of Biomedical Engineering, Atlanta, GA *Emory University School of Medicine, Atlanta, GA Purpose: Increased aortic pulse wave velocity (PWV) results from decreased vessel compliance and can be due to increased age, atherosclerosis and/or hypertension (Farrar et al. Circ. 1991). The goal of this study was to compare reproducibility of PWV measurements obtained via applanation tonometry (AT), a clinically accepted method, to a new PCMR method using cross-correlation analysis. Methods: PWV was measured in seven asymptomatic patients with an elevated CT coronary calcium score, and ten healthy volunteers. Oblique sagittal images in the plane of the aorta (“Candy Cane” view) were acquired with a velocity encoding in the foot-head direction. Cross-correlation was used to determine time shifts between velocity curves at adjacent points along the aorta. The slope of a regression line fit to the transit time vs. location data was used to estimate PWV (Fielden et al. JMRI 2008). For AT, a Sphygmocor device (AtCor Medical) was used to measure pressure waveforms at carotid and femoral locations and the PWV was determined using a standard transfer formula. Inter-scan variability for each method was assessed with the coefficient of variation (CoV). Results: PCMR had a significantly lower inter-scan CoV than AT (Figure 1). The better reproducibility may allow for the enrollment of fewer patients for trials of interventions targeted at vessel wall compliance. Conclusion: PWV estimates using PCMR and cross-correlation processing have Figure 1: Inter-scan coefficient of variation for AT and PCMR superior reproducibility as compared to AT. 126 MR-Angioclub East Lansing 2009 P.9 Motion-compensated, flow-independent, non-contrastenhanced renal MR angiography Gregory J. Wilson1,2, George R. Oliveira2, Liesbeth Geerts1, Jeffrey H. Maki2 1 – Philips Healthcare, Cleveland, USA and Best, Netherlands; 2 -- University of Washington Dept of Radiology, Seattle, WA, USA Purpose Evaluate a new MR technique for high resolution imaging of renal arteries. While breath-hold contrast-enhanced MRA (CE-MRA) of renal arteries provides useful images, the achievable spatial resolution can be limited by breath-hold duration, cardiac motion [1], respiratory motion [2], and the contrast agent bolus profile [3]. The flowindependent, non-contrast-enhanced method implemented here reduces or eliminates these resolution-limiting effects. Methods The technique has been evaluated in 3 healthy volunteers. The cardiac- triggered, balanced TFE acquisition employed respiratory navigator gating, fat saturation, and magnetization preparation (T2 Prep). Data was acquired only during diastole, the most quiescent period of flow and cardiac-related motion, and without inversion pulses. As a result, the sequence was flow-independent. Axial volumes were acquired with resolution of 1.68x1.72x2.4mm3 (reconstructed to 0.73x0.73x1.2mm3) in 4:26 with 70% navigator efficiency and 2 averages. Results The acquisition was successful in each volunteer. High resolution images were produced without the limitations of motion or bolus profile blurring. Flow independence was demonstrated by lack of in-flow signal modulation in coronal acquisitions, and lack of venous signal suppression by application of in-flow saturation bands. As a consequence, veins were visible in the images. Conclusion The new technique provided high resolution images for evaluation of renal artery lumen morphology. Further study is planned to evaluate the diagnostic value of this technique. Figure 1: High resolution, axial source image displaying left renal artery. Figure2: Curved reformat showing bilateral renal arteries. References 1-Kaandorp DW, JMRI2000;12:924-8; 2-Vasbinder GB, JMRI2002;16:685-96; 3-Fain SB, MRM1999;42:1106-16. MR-Angioclub East Lansing 2009 127 P.10 Vascular response during visual stimulation at 3T MRI: functional phase contrast angiography (fPCA) study 1 1 Sang-Hoon Kim , Chang-Ki Kang , Young-Bo Kim1, Zang-Hee Cho1, 2 1. Neuroscience Research Institute, Gachon University of Medicine and Science, Korea 2. Department of Radiological Sciences, University of California, USA Purpose: The purpose of present study is to investigate direct arterial response with quantitative velocity analysis using phase contrast angiography (PCA) sequence [1,2]. Methods: PCA sequence was used with the following parameters: 0.67x0.67x1.0 mm3 of imaging resolution, 36 slices, and 20 cm/s of velocity encoding (VENC) for all directions. Total acquisition time was 10 m 18 s for 3 sessions, including 2 rests and 1 visual stimulation (a flashing checker board). A custom-built surface coil was designed for angiographic purpose and tuned to 3T. Twelve subjects participated in this study. Circular region of interests (ROIs) consisting of 21 pixels was used to measure the velocity difference on the different diameters. We found the ROI showing a maximum change and defined as peak ROI and distal ROIs for vessels distal to the peak ROI. Velocity changes of the selected ROIs were measured. Results: Fig. 1 showed the vascular change acquired from PCA and calculated the velocity and its percentage change during stimulation. Velocity change during stimulation was clearly delineated, which were indicated with arrows (a). Mean velocity change between rest and stimulation periods were 0.64 and 0.42 cm/s from a peak ROI and distal ROIs (b), respectively, which are corresponding to 42.3 and 29.6 of percentage change as shown in Fig. 1(c). Conclusion: PCA for functional angiography (fPCA) at 3T MRI could provide vascular response during brain function, including the vasculature as well as the velocity changes of related vessels. References: 1. Iadecola C, et al., J Neurophysiol 1997. 2. Cho ZH, et al., Neuroimage 2008. Fig. 1. Quantitative velocity analysis of fPCA 128 MR-Angioclub East Lansing 2009 P.11 Optimization of Phase-contrast MR-based Flow Velocimetry and Shear Stress Measurement Taeho Kim1, Jihyea Seo1, Seongsik Bang1, Hyeonwoo Choi1, Yongmin Chang2, Tae-Seob Chung3, Jongmin Lee4 1 2 4 Department of Biomedical Engineering, Molecular Medicine, and Radiology, Kyungpook National 3 University; Department of Radiology, Yonsei University Purpose: To measure the pixel-by-pixel flow velocity and shear stress from PC MR images, an optimized method was suggested and verified. Material and Methods: A self-developed straight steady flow model was scanned by 3T MRI with a velocity-encoded PC sequence. The pixel-by-pixel flow velocity and shear stress was measured using self-made program and were compared with the physically measured reference data. A comparison between the program and a commercial velocimetry system was also performed. Subsequently velocity and shear stress were measured in curved steady flow model to confirm shifted peak velocity and shear stress toward the outer side of lumen. Results: The velocity measured with self-made program showed a significant correlation with the physically measured velocity and was superior to the commercial system (R2 = 0.85 vs. 0.75, respectively). The calculated mean shear stress had a significant correlation with the physically measured mean shear stress (R2 = 0.95). The curved flow model showed a significantly shifted peak velocity and shear stress zones toward the outside of the flow. Conclusion: The technique to measure pixel-by-pixel velocity and shear stress of steady flow from velocity-encoded phase-contrast MRI was optimized and verified to be superior to the use of a commercial system. MR-Angioclub East Lansing 2009 129 P.12 Blood Flow Patterns in the Abdominal Aorta of Mice: Implications for AAA Localization 1 1 3 2,3 1,2 Smbat Amirbekian , Robert Long , Michelle Consolini , W. Robert Taylor , John Oshinski Department of Radiology1, Biomedical Engineering2, and Medicine (Cardiology)3, Emory University School of Medicine, Atlanta, GA Introduction: Abdominal aortic aneurysms (AAA) localize almost exclusively in the infrarenal aorta in man. Humans experience a period of reverse flow during early diastole in the infra-renal aorta during each cardiac cycle. This reverse flow causes oscillatory wall shear stress (OWSS) to be present in the infra-renal aorta of humans and has been linked to AAA formation. In contrast to humans, all mouse models of AAA form aneurysms in the supra-renal aorta. The presence of reverse flow in the mouse aorta is unknown. Purpose: The goal of this study was to examine flow patterns in the mouse aorta and evaluate whether reverse flow exist in the infra-renal or supra-renal aorta. Methods: Blood flow was measured with a phase contrast magnetic resonance (PCMR) sequence the supra-renal and infra-renal abdominal aorta of 18 wild type C57BL/6 mice and 15 ApoE-/- mice (most prevalent mouse model of AAA). Measurements were made using a 4.7T Varian system. Results: Results indicate that unlike humans, there is no reversal of flow in the infra-renal or supra-renal aorta of wild type or ApoE-/- mice. Distensibility of the mouse aortic wall as a percent of diameter (63 + 18%), is higher than reported values for the human aorta. Conclusion. Wild type and ApoE-/- mice do not experience the reverse flow associated with OWSS and AAA in the infra-renal aorta that is observed in humans. 130 MR-Angioclub East Lansing 2009 P.13 3D Vessel Wall imaging of multiple vascular beds 1 Niranjan Balu , Jinnan Wang2, Chun Yuan1 1. University of Washington, Seattle, WA 2. Philips Medical Systems, BriarCliff Manor, Ny Introduction: Black-blood MRI of atherosclerotic plaque allows identification of plaque morphological and compositional features and thereby helps in assessing disease burden and risk of thromboembolism due to plaque Representative rupture. Recently, a fast isotropic 3D black-blood images (a) b sequence for carotids has been shown to provide Sagittal carotid with calificate comprehensive plaque information in all spatial (b) Montage directions [1]. Since atherosclerosis is a systemic from slices to show entire disease, plaque is coexistent in multiple vascular femoral artery beds. 3D MRI of multiple vascular beds is an ideal covered (c) tool to study the pathophysiology of Cross-section of abdominal atherosclerosis as a systemic disease. aorta Aim: To develop a time-efficient 3D isotropic imaging protocol for carotid vessel wall imaging of a the carotids, aorta and femoral arteries Methods: A 3D FLASH sequence with MSDE [2] preparation was optimized (table) for carotid, aorta and femoral imaging. Results: Vessel wall images of diagnostic image quality were obtained in the carotids, aorta and femoral arteries (figure). Scan time was approximately 2 min, 3.5 min and 10 min for coverage of approximately 15 cm, 16 cm and 50 cm in the carotids, femoral and aorta respectively. Conculsions: Time-efficient vessel wall imaging c of large arteries can be achieved within a one-hour scan time by use of the new sequence. References: [1] Balu N et al, ISMRM 09 [2] Wang et al, MRM 07;58:973-981 Vascular Bed Acquisition plane Resolution, mm2 Field-Of-View, mm2 Slice thickness , mm # of slices TR/TE, ms Flip angle, ° Turbo factor Bandwidth, Hz/pixel Number of averages Number of stations Scan time, min:s Carotid Coronal 0.7 250×160 0.7 100 10/4.8 6 90 134.3 1 1 2:03 Aorta Coronal 1.4 350×300 1.4 86 7.7/3.7 6 35 269.1 1 1 3.23 Femoral Coronal 1.0 300×350 1.0 150 7.6/3.5 6 100 191.6 1 3 10.24 MR-Angioclub East Lansing 2009 131 P.14 Feasibility Study of Combining 3D SSFP with T2prep Inversion Recovery (T2IR) for Black Blood Vessel Wall Imaging Keigo Kawaji1,2, Thanh Nguyen2, Pascal Spincemaille2, Martin R. Prince2, Yi Wang1,2 1 Department of Biomedical Engineering, Cornell University, Ithaca, NY. 2 Department of Radiology, Weill Cornell Medical College, New York, NY. Background: T2IR preparation vessel wall imaging provides flow-insensitive global black-blood (BB) suppression at the expense of SNR [1-3]. A recent study [3] has demonstrated the feasibility of T2IR preparation in 2D Fast-Spin-Echo for BB vessel wall imaging. In this report, we examine the feasibility of imaging vessel walls using a 3D Steady-State Free Precession (SSFP) sequence with T2IR preparation. Methods: 4 healthy volunteers were scanned using a 4-channel cardiac coil for signal reception. Axial views of the popliteal artery were acquired using gated 3D SSFP with T2IR preparation on a 1.5T scanner (GE Signa HDx) using the following scan parameters: imaging matrix 256x256x50, slice thickness=2mm, flip angle=60o, FOV=16cm, NEX=2, VPS=100, TE=1.6ms, TR=4.0ms, TEeff=120ms, TI=225-375ms, with a scan time of approximately 4 mins for a nominal heart rate of 60bpm. Fat saturation was also used. Results: a ) c) Figures 1 a,b) show axial views of the popliteal artery, and 1 c) shows a reformatted view from the 3D T2IR SSFP images. The vessel walls were clearly depicted in these images. Discussion: b ) Preliminary data demonstrate the feasibility of BB imaging of the vessel wall using a 3D T2IR prepared SSFP sequence. References: 1. Brittain et al. MRM 1997; 2. Liu et al. ISMRM 08 pp3079; 3. Nguyen et al. ISMRM 09 pp607. 132 Figure 1: a) Axial view of 3D T2IR SSFP image of knee. b) Popliteal artery c) Reformatted view. MR-Angioclub East Lansing 2009 P.15 Identification of Optimal First-Order Gradient Moment for Flow-Sensitive Dephasing (FSD) Preparation Z. Fan1,2, X. Bi3, and D. Li1,2 Radiology, 2Biomedical Engineering, Northwestern University, Chicago, IL, USA; 3 Siemens Medical Solutions USA, Inc., Chicago, IL, USA 1 Purpose: The first-order gradient moment, m1, is a measure of the blood signalsuppression capability of the FSD preparative module. This work aimed to develop an m1scout approach to rapidly identify the optimal m1, which may help improve vessel wall imaging and FSD-based MRA quality. Methods: FSD-induced signal suppression is voxel size-dependent as its underline mechanism is the intravoxel velocity variation [1]. For a FSD-prepared 3D isotropic resolution imaging with major flows in the readout direction, we proposed to run the same imaging seqence in 2D mode to rapidly identify the optimal m1. This requires the 2D imaging plane perpendicular to the major vessel of interest, FSD gradient pulses applied in the slice-select direction, and the in-plane resolution identical to that of 3D imaging. We hypothesized that the flow-void effects to occur in the 3D case could be reflected by the 2D case with the same m1 used. This approach was tested on flow phantom (Gd-doped water 0.25mM, a silicone tube of 4.8-mm ID connected to a pump) at various flow rate (15, 20, 30, 40, 50, 60 cm/s). The 2D m1-scout sequence was implemented based on FSDprepared balanced SSFP which was also used in 3D imaging for verification. Spatial resolution: isotropic 0.9-mm for 3D, 0.9-mm in-plane and 5-mm thickness for 2D. The 2D scan continuously acquired 11 measurements (images) (Fig. 1), whereas the 3D scan acquired only 6 images due to long scan time. Results: The lumen signal intensity measurements from the 2D and 3D images were significantly correlated at all the velocities tested (Mean Pearson correlation coefficient = 0.988, p < 0.001). Fig. 2 shows the case of flow rate = 20 cm/s. Conclusion: The optimal m1 value may be identified using the fast 2D scout approach. In vivo experiments are underway to further verify its applicability. References: [1] Nguyen TD et al. JMRI 2008; 28:1092-100. Fig 1. Schematic of the 2D m1-scout scan. A total 11 images were collected. The first uses m1 = 0, while the latter uses incremental m1 values (user specified.) Fig 2. Corresponding 2D and 3D images (a) and signal intensity measurements (b). 2D 3D a 60 80 100 120 140 1.20 Signal Intensity m1= 0 2D Scout 1.00 0.80 V = 20 cm/m 0.60 3D Scan 0.40 0.20 0.00 b 0 60 70 80 90 100 110 120 130 140 150 First-order gradient moment (mT.ms2/m) MR-Angioclub East Lansing 2009 133 P.16 Combined Segmentation of Lumen and Intraplaque Hemorrhage in Black-blood T1-Weighted Carotid Imaging Rahul Sarkar1, General Leung1,2 and Alan R. Moody1,2 University of Toronto and 2Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada 1 Purpose: We investigated the use of a novel automatic segmentation approach to allow localization of intraplaque hemorrhage (IPH) with respect to the lumen in black-blood T1 weighted carotid images. Methods: From the original image, a fuzzy connectedness map is produced using a generalized fuzzy spel affinity [1] for lumen detection. The fuzzy thresholded lumen contour is downsampled to form control points for a b-spline curve to smooth image noise at the boundary. A radial distance metric from the segmented lumen center of mass is combined with intensity thresholding for segmentation of IPH. Images used for training the fuzzy spel affinity and testing were obtained as previously described [2]. Results: Figure 1 shows the segmentation results produced from a representative image showing intraplaque hemorrhage. The fuzzy segmentation step is highly efficient at gross lumen detection, but demonstrates sensitivity to noise along border pixels. The derived bspline curve is effective in smoothing the border noise without significantly distorting the lumen shape. IPH is accurately detected by thresholding within a defined radial distance metric (5mm) from the segmented lumen. Conclusion: The described approach is effective in automatically segmenting the lumen and IPH to allow their co-localization in black blood T1weighted images. Figure 1. From left: 1) Original image; 2) Fuzzy detected lumen (red) showing border noise (blue arrow); 3) b-spline corrected fuzzy lumen (red) showing reduced border noise (green arrow); 4) Corrected lumen (red) and automatically segmented IPH (yellow). References: [1] Carvalho PAA 1999. [2] Leung ISMRM 2009. 134 MR-Angioclub East Lansing 2009 P.17 Automatic Registration of Multiparametric T1 Weighted Images Using FOV-Selective Mutual Information Rahul Sarkar1, General Leung1,2 and Alan R. Moody1,2 University of Toronto and 2Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada 1 Purpose: Multiparametric T1weighted imaging may be useful in identifying several features of atherosclerotic plaque [1]. Here we investigate a technique to automatically and reliably register the parametric images using normalized mutual information [2]. Methods: All algorithm and user interface development was done in MATLAB (The Mathworks). Mutiparametric datasets were obtained as previously described [1]. For each dataset, a normalized mutual information algorithm was applied to register blackblood post-contrast and bright-blood delayed enhancement images to the reference blackblood pre-contrast image. Algorithm performance was observed under different field of view (FOV) selections for reference and non-reference images. Results: The algorithm was effective in accurately registering images of various nonreference image FOVs (Figure 1). Reference image FOV was an important factor in algorithm efficiency. Small reference FOVs (40mm x 40 mm) including the vessel wall and few surrounding features proved optimal in terms of accuracy and reduced computational time significantly. Deformable motion effects in large reference FOVs sometimes led to misregistration that was corrected by reducing the FOV size. Conclusion: Multiparametric T1 weighted images may be accurately co-registered with a normalized mutual information algorithm using small reference image FOVs. Figure 1. Multiparametric T1 images with pre-contrast black blood (left), post-contrast black blood (middle), and bright blood delayed enhancement (right). Top panel shows unregistered images of different FOVs. Bottom panel shows the same images registered to the reference pre-contrast image. References: [1] Leung ISMRM 2009. [2] Collignon IPMI 1995. MR-Angioclub East Lansing 2009 135 P.18 MR fluid dynamics using 4D-Flow for intracranial aneurysms with growing blebs and a ruptured intracranial aneurysm Haruo Isoda1, Hisaya Hiramatsu2, Yasuhide Ohkura3, Takashi Kosugi3, Masaya Hirano4, Shuhei Yamashita1, Yasuo Takehara1, Marcus T. Alley5, Harumi Sakahara1; 1Department of Radiology and 2 Department of Neurosurgery, Hamamatsu University School of Medicine, Hamamatsu, Japan; 3 4 Renaissance of Technology Corporation, Hamamatsu, Japan; GE Yokogawa Medical Systems, Hino, Japan; 5 Department of Radiology, Stanford University School of Medicine, Stanford, USA. PURPOSE: The purpose of our study was to investigate the relationship between hemodynamics and growing blebs or intracranial aneurysm ruptures with the use of MR fluid dynamics (MRFD) using 4D-Flow (1). METHODS: This study included an unruptured 6mm right IC-Ach artery aneurysm with growing bleb, an unruptured 4mm left IC aneurysm (C2 and C3 segment) with growing bleb and a ruptured 8mm right IC-Ach artery aneurysm with hematoma in the medial aspect of the right temporal lobe. 4D-Flow was performed by a 1.5T MR scanner. We calculated and evaluated streamlines, wall shear stress (WSS) and the oscillatory shear index (OSI) of these intracranial aneurysms based on 4D-Flow with our in-house software. RESULTS: The top of the spiral flow was near the bleb in the unruptured IC-Ach artery aneurysm. The WSS was lower and the OSI was higher near the bleb. The top of the spiral flow was present at the growing bleb of the unruptured IC artery aneurysm and the WSS was lower and the OSI was higher at this point. A disturbed spiral flow was noted in the ruptured IC-Ach artery aneurysm. Lower WSS and higher OSI were observed in an area containing the rupture point. CONCLUSION: Bleb formation or rupture points were observed near the apex of the spiral flow with low WSS and high OSI. Based on previously published papers reporting that low WSS or high OSI causes degeneration and apoptosis of endothelial cells, we hypothesized that endothelial cells at the apex of the spiral flow in the aneurysms were much more vulnerable to degeneration or apoptosis. REFERENCES: 1. Markl M, et al. J Magn Reson Imaging. 2003;17:499-506. 136 MR-Angioclub East Lansing 2009 P.19 Phase-Field Dithering for Active Catheter Tracking 1 Charles L. Dumoulin , Richard P. Mallozzi2, Robert D. Darrow3, and Ehud J. Schmidt4 1 2 3 Cincinnati Children’s Hospital, ONI Medical Systems, Inc., General Electric Global Research, and 4 Brigham and Women’s Hospital Purpose: Active MR tracking can become difficult in low Signal-to-Noise (SNR) conditions, especially when magnetic susceptibility of the device is not well matched to its surroundings and when unwanted MR signal is coupled into the receive channel (e.g. from poorly shielded cables). A strategy to increase the robustness of active MR tracking of micro-coils in these conditions was developed, and tested. Methods: The method employs dephasing magnetic field gradient pulses that are applied orthogonal to the frequency encoding gradient pulse used in conventional point-source MR tracking. In subsequent acquisitions the orthogonal dephasing gradient pulse is rotated, while maintaining a perpendicular orientation with respect to the frequency encoding gradient. The dephasing gradient creates a spatially-dependent phase shift in directions perpendicular to the frequency encoding gradient. Since the desired MR signal for robust MR tracking comes from the small volume of nuclear spins near the small detection coil, the desired signal is not dramatically altered by the dephasing gradient. Undesired MR signals arising from larger volumes (e.g. due to coupling with the body and/or surface coils), on the other hand, are dephased and reduced in signal intensity. Since the approach requires no a priori knowledge of the micro-coil orientation with respect to the main magnetic field, data from several orthogonal dephasing gradients is acquired and analyzed in real-time. One of several selection algorithms is employed to identify the “best” data for use in coil localization. Results and Conclusions: This approach was tested in flow phantoms and animal models, with several multiplexing schemes, including the Hadamard and zero-phase reference approaches. It was found to provide improved MR tracking of un-tuned microcoils. It also dramatically improved MR tracking robustness in low SNR conditions and permitted tracking of micro-coils that were inductively coupled to the body coil. MR-Angioclub East Lansing 2009 137 P.20 Interventional Device Tracking and Imaging Using an Extensible Real-Time System Ethan Brodsky and Orhan Unal Radiology and Medical Physics, University of Wisconsin, Madison, WI, USA PURPOSE: To demonstrate tracking of interventional devices using RTHawk, extensible software architecture for real-time MR imaging. METHODS: RTHawk is a software architecture that permits a variety of pulse sequences, acquisition trajectories, and reconstruction techniques to be easily developed and interleaved at run-time on GE scanners [1]. We have previously demonstrated simultaneously tracking a catheter and imaging an endovascular device using a customdesigned software solution that was tied to a particular pulse sequence implementation and was very difficult to modify or extend [2,3]. We have therefore implemented two catheter tracking techniques in RTHawk – the previously described 3DPR technique that is amenable to simultaneous tracking and imaging, as well as a simple three-orthogonalview tracking acquisition that can easily be interleaved with Cartesian imaging. RESULTS AND DISCUSSION: Initial testing shows that the previously developed tracking and imaging techniques can be easily adapted to run in this environment, without requiring any pulse sequence programming. A real-time display shows one-dimensional profiles for each axis from which the signal peak is detected and tracked. CONCLUSIONS: Initial results suggest that interventional sequences can easily be adapted to work within this framework. It abstracts away much of the complexity of pulse sequence and reconstruction programming, allowing simple intuitive representations of the acquisition trajectory and reconstruction pipeline. Visualization of final and intermediate results is also greatly simplified. Figure 1: These one-dimensional profiles show the signal picked up by the endovascular device’s tip tracking coil for projections along each of the orthogonal gradient axes. Pulse sequences was Fast GRE, 4° flip, 12.5 ms TR, 256 sample readout over a 40 cm FOV for a 1.6 mm resolution. Figure 2: This plot of peak position over time shows the tip position as the device is moved in a circle in the x-z plane. REFERENCES: [1] Santos et al. Proc IEEE Eng Med Biol Soc. 2:1048,’04 [2] Brodsky et al. MRM 56:247, ‘06 [3] Unal et al. Proc MRA Club 18:4.2, ’06 138 MR-Angioclub East Lansing 2009 P.21 Simplified Catheter-based Multimode Coil for Active MR Tracking and Intravascular Imaging Mahdi Salmani Rahimi, Krishna Kurpad, Orhan Unal Departments of Biomedical Engineering, Radiology and Medical Physics, University of WisconsinMadison PURPOSE: To simplify the design of a multi-mode intravascular MR catheter-based coil that combines the functionalities of an active tip-tracking coil and an imaging coil and to resolve the ambiguity of having two tip-tracking signal peaks. METHODS: Unlike earlier designs [1] that had a separate solenoid at the catheter tip to provide the localized tracking signal, in this design the inductor in the required matching network of the tuned imaging coil has been exploited for tip tracking. Quarter-wavelength π-matching network was placed a at the distal tip with a wirewound solenoid as the inductor. 1a shows the coil schematic. Cable lengths were c Tracking Coil Decoupling Box and GE 1.5T Scanner Amp. Figure b Imaging Loop (λ/4 long) adjusted so that the solenoid in the matching network is in resonance during the receive cycle of the scan. To increase the number of turns in the inductor to improve tip-tracking signal and avoid λ/4 Transformer Freq. Figure 1: (a) Schematic of the catheter coil design. (b) 313 micron resolution axial image of a grapefruit obtained with the imaging loop using a 2D GRE technique. (c) Signal profile from manual prescan, TG=100. Tracking coil peak is visible and easily detectable from broader imaging signal in the coil with one counter turn. unwanted increase in the competing imaging signal, some turns in the imaging coil have been wound in the opposite direction. RESULTS: Figure 1b shows a high resolution axial image of a grapefruit obtained with the imaging loop of the multimode probe using a 2D GRE sequence (FOV=80x80mm, 256X256, TR/TE=200/3.1 ms, FA=90o, and NEX=5). Since there is only one solenoid in this design, the ambiguity in determining tip position caused by having more than one localized peak in earlier designs can also be avoided as seen in Figure 1c. CONCLUSIONS: Our preliminary results suggest that multimode intravascular probes can be simplified while improving active tip-tracking and imaging functionalities. REFERENCES: [1] Unal O, et al. ISMRM 1398, 2006. MR-Angioclub East Lansing 2009 139 P.22 Transmit Power Optimization for Tracking, Wireless Marker and Imaging applications of a Multi-mode Endovascular coil. Krishna N. Kurpad and Orhan Unal Departments of Radiology and Medical Physics, University of Wisconsin-Madison PURPOSE: To optimize body coil transmit power settings to manipulate transverse magnetization in various regions in the vicinity of a MR multi-mode endovascular coil for optimal operation in tip tracking, wireless marker, and endovascular imaging modes. METHODS: The multimode coil 1,2 (Fig. 1a), being inductively coupled to the transmit coil, behaves as a local B1 field magnifier. The seriesconnected tip tracking and imaging components of the multimode coil are constructed with different conductor densities, resulting in different transverse magnetization in their vicinity. A series of 21 coronal images of a tube phantom containing saline solution doped with copper sulphate (T1=150ms) was obtained using a FGRE pulse sequence (TR/TE= 200/4 ms, FA=90o, FOV=6cm, 256x256) on a GE Signa 1.5T clinical scanner. The images were obtained by incrementing the transmit gain setting in steps of 1 dB between the limits of 0 dB and 20 dB. Graphs of the mean signal intensity in three small regions (Fig. 1b) located close to 1) the tracking coil, 2) the conductors of the imaging coil, and 3) one cm away from the imaging coil were plotted against the transmit gain setting for the corresponding images. RESULTS: The graphs in Fig. 2 show that transmit power settings may be suitably adjusted to generate 1) distinct tracking peak, 2) broad region of sensitivity for imaging, or 3) visualize the conductors of the multimode coil for wireless marker applications. CONCLUSIONS: Preliminary results show that the current induced in the multimode coil by the transmitted B1 field results in magnetization varying due to transverse the varying conductor density of its various components. The transmit power may therefore be optimized for optimal operation of the multimode coil in all three functional modes. REFERENCES:1Kurpad KN,et al, ISMRM 292, 2007, 2Unal O, et al. ISMRM 1398, 2006 140 MR-Angioclub East Lansing 2009 Sponsors Lantheus Medical Imaging, a worldwide leader in diagnostic imaging for 50 years, is committed to developing innovative solutions for the diagnosis and management of disease. For more information, please visit www.lantheus.com. "Bayer Schering Pharma AG has a worldwide leading position in developing and manufacturing contrast media for use in x-rays, computed tomography (CT), magnetic resonance imaging (MRI) and in contrast media application systems. www.bayerhealthcare.com Bracco Imaging S.p.A, is one of the World’s leading companies in the diagnostic imaging business. Bracco Imaging S.p.A products are distributed in USA by Bracco Diagnostic Inc., Princeton. www.braccoimaging.com GE Healthcare's expertise in medical imaging and information technologies, medical diagnostics, patient monitoring systems, disease research, drug discovery and biopharmaceuticals is dedicated to detecting disease earlier and tailoring treatment for individual patients. For more information visit www.gehealthcare.com syngo TimCT offers continuous table move for CT-like scanning. Setting the trend in MRI with MAGNETOM ESSENZA, Avanto, Espree, and Verio. Siemens. Answering the questions of life. www.siemens.com/mr Guerbet is the only pharmaceutical group fully dedicated to medical imaging, offering a comprehensive range of contrast agents for X-ray and MRI worldwide. Guerbet’s innovation focuses on molecular imaging for MRI or nuclear medicine applications, enabling diseases to be targeted earlier and more precisely. www.guerbet.com MR-Angioclub East Lansing 2009 141 Sponsors Philips Healthcare is a premier supplier of healthcare solutions. The MR portfolio provides unique, advanced features enabling our users to extend the realms of vascular MR imaging. www.medical.philips.com TopSpins Inc. is dedicated to helping MR imaging centers implement safe, accurate, state-of-the-art contrast enhanced MR Angiography and dynamic contrast-enhanced MR imaging studies for the benefit of their patients. www.topspins.com Robarts Research Institute harnesses the power of cell biology, genomics and advanced imaging technologies to investigate disorders of the neurological, cardiovascular and immune systems. It is a research Institute within the Schulich School of Medicine & Dentistry at The University of Western Ontario. www.robarts.ca Founded in 1878, The University of Western Ontario is one of Canada's oldest universities and today is committed to providing the best student experience among Canada's leading researchintensive universities. A vibrant centre of learning located in London, Ontario, Western is home to more than 30,000 students and attracts external support of over $200 million annually for research projects. www.uwo.ca As an integral part of the Michigan State University HealthTeam, the Department of Radiology has served the Mid-Michigan community since 1975. Our dedicated team of health care professionals has pioneered research and clinical applications in radiology. www.rad.msu.edu 142 MR-Angioclub East Lansing 2009 Insert Lantheus file MR-Angioclub East Lansing 2009 143 Notes 144 MR-Angioclub East Lansing 2009 Notes MR-Angioclub East Lansing 2009 145 Notes 146 MR-Angioclub East Lansing 2009 Notes MR-Angioclub East Lansing 2009 147 Presenter Index Page A Abela, George The Role of Cholesterol Crystals in Acute Cardiovascular Events: Indentifying the Cause for Gender Differences in Clinical Presentation Admirral - Behloul, Faiza Hybrid of Opposite Contrast MR Angiography of the Brain Artz, Nathan Assessing Kidney Perfusion using Arterial Spin Labeling and Radial Acquisition for Rapid Characterization of Inflow Dynamics Aschauer, Manuela Gadofoveset Excretion into Human Breast Milk CE-MRA with tailored 3D random sampling patterns and nonlinear parallel imaging reconstruction (P1) 118 73 101 54 119 B Balu, Niranjan 3D Vessel Wall imaging of multiple vascular beds (P13) Barnes, Samuel High Resolution Simultaneous Angiography and Venography (MRAV) with a Single Echo Bernstein, Matt Multicenter Studies: Lessons Learned from ADNI Bhat, Himanshu Contrast-Enhanced Whole-Heart Coronary MRA at 3T Using Gradient Echo Interleaved EPI (GRE-EPI) Blackham, Kristine Robust Clinical Application of Time-Resolved MRA(P7) Bley, Thorsten Non-invasive Trans-Stenotic Pressure Measurements with 3D Phase Contrast MRA: Validation against Endovascular Pressure Measurements in Swine Block, Walter Imaging Capabilities for Real-time Guidance and Verification of Transcatheter Arterial Chemoembolization (TACE) Procedures Bousell, Loic 4D time-resolved MR angiography for non-invasive pulmonary postembolization AVM patency assessment Brodsky, Ethan Interventional Device Tracking and Imaging Using an Extensible Real-time System (P20) 131 76 114 90 125 61 102 83 138 C Carr, James Non Contrast MRA of the Hand in Patients with Raynauds disease using Flow Sensitized Dephasing Prepared SSFP Carroll, Timothy Radial Sliding Window MRA in Pulmonary Hypertension Choi, Grace MRA with the “No Phase Wrap” Chung, Tae-Sub Obstruction of IJV by Asymmetry of Lateral Mass of Atlas on Head and Neck CEMRA and Contrast CT 148 MR-Angioclub East Lansing 2009 50 84 86 78 Author Index Page D Dumoulin, Charles Phase-Field Dithering for Active Catheter Tracking (P19) 137 E El-Boubbou, Kheireddine Targeted Glyco-Magnetic Fe304 Nanoprobes for Detection and Molecular Imaging of Atherosclerosis Essig, Marco Intraindividual comparison between multislice CT and 4D TWIST MRA in the assessment of residual cerebral ateriovenous malformations – a prospective study protocol 110 71 F Fan, Zhaoyang 3D Noncontrast MRA using FSD – Prepared Balanced SSFP Identification of Optimal First-Order Gradient Moment for Slow-Sensitive Dephasing (FSD) Preparation (P15) Fielden, Samuel Balanced-gradient TSE for Non-Contrast Peripheral MRA Francois, Christopher Flow assessment of arterial dissections using 3D radial phase contrast MR Angiography Frydrychowicz, Alex Analysis of aortic hemodynamics after treatments for coarctation using flow-sensitive 4D MRA at 3T 49 133 27 63 62 G Ge, Lan Myocardial Perfusion MRI in Canines with Improved Spatial Coverage, Resolution and SNR Geerts, Liesbeth Non-CE Imaging of the Pulmonary Arteries Goldfarb, James Cardiac Imaging: Methods for the Detection of Intramyocardial Fat Grabow, Ben Temporal Filtering for Sliding Window Time-resolved Angiography; Beyond Density Compensation Solutions Grist, Thomas Overview of Gd-BOPTA Phase III Trial for CEMRA: What are the results, and how do we move forward? Griswold, Mark A Simple View of Compressed Sensing and How it Could Change Everything We Do in MRI and MRA 32 24 92 94 55 29 H Haacke, Mark High Resolution Perfusion Weighted Imaging 66 Susceptibility mapping as a means to image veins 104 MR-Angioclub East Lansing 2009 149 Author Index H Hadley, Rock A 16 Channel Anterior Neck RF Coil for Cervical Carotid MRA Haider, Clifton A Comparison of Time-Resolved 3D CE-MRA with Peripheral Run-off CTA in the Calves Hibberd, Mark An Update on the Clinical Experience with Gadofosveset Hu, Peng Non-Contrast Enhanced Pulmonary Vein MRI with a Spatially Selective Slab Inversion Preparation Hulet, Jordan Improved Carotid Imaging with HASTE using a reduced FOV and increased gradient performance (P3) Page 42 47 52 85 121 I Igase, Keiji Our Strategy for the Surgical Planning with 3T MRA in Detecting Unruptured Cerebral Aneurysms Isodo, Haruo MR fluid dynamics using 4D-Flow for intracranial aneurysms with growing blebs and a rupture intracranial aneurysm (P18) 72 136 J Jeong, Hyun CAMERA: Contrast-enhanced Angiography with Multi-Echo and Radial kspace Johnson, Casey Two-Station Time-Resolved CE-MRA of the Lower Legs Johnson, Kevin Accelerated Time Resolved Inflow with 3D Radial bSSFP Angiographic and Hemodynamic Assessment of the Hepatic Vasculature in Portal Venous Hypertension using High Resolution PC VIPR 106 51 25 95 K Kang, Chang-Ki Vascular response during visual stimulation at 3T MRI: functional phase contrast angiography (fPCA) study (P10) Kawaji, Keigo Feasibility Study of Combining 3D SSFP with T2prep inversion Recovery (T2IR) for Black Blood Vessel Wall Imaging (P14) Kecskemeti, Steven Stack of Stars 4D Phase Contrast Velocimetry of the Circle of Willis Kerwin, William Fibrous Cap Thickness Assessment: Fact or Fiction? Kim, Bum-soo Low Dose 3D Time-Resolved MR Angiography of the Supraaortic Artery: Correlation to High Spatial Resolution 3D Contrast-Enhanced MRA Kim, Seong-Eun Improved Black Blood Multi-Contrast Protocol for In-vivo Atherosclerotic Imaging 150 MR-Angioclub East Lansing 2009 128 132 65 38 77 40 Author Index K Page Kurpad, Krishna Transmit Power Optimization for Tracking, Wireless Marker and Imaging applications of a Multi-mode Endovascular coil (P22) 140 Page L Ladd, Mark Toward Abdominal MRA at 7 Tesla Latourette, Matthews R*2 Calibration Phantoms for Cardiovascular Studies (P5) Laub, Gerhard Low-Dose 4D MR Angiography Lauzon, Louis Non Contrast-Enhanced MR identification of DVT Lee, Jongmin Optimization of Phase-Contrast MR-based Flow Velocimetry and Shear Stress Measurement (P11) Leiner, Tim Gadobenate dimeglumine vs. gadopentetate dimeglumine for peripheral MR angiography: Comparison with DSA Li, Rui Gradient Echo Based Sequence Provides More Information from Ex Vivo Carotid Plaque Specimens Liu, Garry Ultrasound guided cardiac gating for coronary MRA Liu, Jing Self-gated Free Breathing 3D Cardiac Cine Imaging with Data Acquisition During Slice Encoding Lu, Zheng-Rong Manganese Based Biodegradable Macromolecular MRI Contrast Agents for Cardiovascular Imaging 116 123 26 103 129 43 39 89 28 60 M Maki, Jeffrey Dose Comparison between Conventional and High Relaxivity Contrast Agents in Peripheral MRA Marinelli, Luca Accelerated velocity imaging using compressed sensing Markl, Michael Wall Shear Stress in Normal and Atherosclerotic Carotid Arteries McNally, Scott MR imaging and significance of flow reversal and carotid atherosclerosis: Initial results Mendes, Jason Handling Motion in Sparse MRA with Whiskers (P2) Miki, Hitoshi Unruptured Intracranial Aneurysms; Detection and Follow-up on 3.0T MRA Mistretta, Chuck 4D DSA and Fluoroscopy: A New Challenge for MRA? MR-Angioclub East Lansing 2009 44 67 68 69 120 87 111112 151 Author Index Page M Mostardi, Petrice Modified CAPR MRA: Improved Imaging of the Arterial and Venous Phases 105 Page N Newman, Tiffany Magnetic Resonance Angiography of the skin for perforator-based autologous breast reconstruction 98 O Oshinski, John Blood Flow Patterns in the Abdominal Aorta of Mice: Implications for AAA Localization (P12) Ota, Hideki Carotid Intraplaque Hemorrhage is Associated with Enlargement of Lipidrich Necrotic Core and Plaque Volume Over TimeL In Vivo 3T MRI Prospective Study 130 35 P Parienty, Isabelle Time-SLIP versus DSA in Patients with Renal Artery Stenosis Parsons, Edward A Re-analysis of MS-325 (gadofoveset trisodium) Clinical Trial Date in Support of US-FDA Approval Peters, Dana 3D spiral high- resolution late gadolinium enhancement Prince,Martin Risk Factors for NSF: A Meta-analysis 99 53 93 58 Q Quick, Harald 7 Tesla Cardiac MRA in Humans 117 R Rahimi, Mahdi Salmani Simplified Catheter-based Multimode Coil for Active MR Tracking and Intravascular Imaging (P21) Reeder, Scott High Temporal and High Spatial Resolution Perfusion Imaging of Hepatocellular Carcinoma of the Liver Robson, Philip Time-Resolved, Vessel-selective, Cerebral Angiography Using Arterial Spin Labeling Roditi, Giles Retrospective 7 year study of the Incidence of Nephrogenic System Fibrosis in Patients Investigated with Gadolinium Contrast-Enhanced Renal Magnetic Resonance Angiography Pictorial Review of Supra-Aortic Artery Pathologies as Visualised with MRA using Blood Pool Contrast Agent (P6) 152 MR-Angioclub East Lansing 2009 139 64 107 57 124 Author Index S Saloner, David Imaging Considerations in Serial Studies of Vascular Disease Saranathan, Manojkumar FINESS (Flow Inversion-prepared Non-contrast Enchancement in the Steady State): A novel technique for non-contrast renal MRA Sarkar, Rahul Combined Segmentation of Lumen and Intraplaque Hemorrhage in Blackblood T1-weighted Carotid Imaging (P16) 115 97 134 Automatic Registration of Multiparametic T1 Weighted Images using FOVSelective Mutual Information (P17) Schiebler, Mark Pulmonary MRA in 75 patients with dyspnea 135 Origin and Frequency of artifacts in Contrast Enhanced Pulmonary MRA in 80 patients with dyspnea Schneider, Guenther Safety of Gadobenate dimeglumine (Gd-BOPTA) in Cardiovascular Imaging of Pediatric Patients 81 Renal MR angiography: multicenter intraindividual comparison of gadobenate dimeglumine and gadofosveset trisodium Seiberlich, Nicole Reconstruction of MR Angiography Images Using Gradient Descent with Sparsification Shah, Dipan Evaluation of Gd-DOTA (DOTAREM) enhanced MRA compared to time-offlight MRA in the diagnosis of clinically significant non-coronary arterial disease at 1.5 and 3.0 Tesla Sheehan, John Flow and Motion-Insensitive Unenhanced MR Angiography of the Peripheral Vascular System – A pilot study in the lower extremity Stafford, Randall Towards Continuously Moving Table MCE Peripheral MRA (P4) Stein, Paul Gadolinium Enhanced Magnetic Resonance Angiography for Pulmonary Embolism: Results of PIOPED III Steinman, David Quantifying Lumen Geometry from Routine Carotid CEMRA Suever, Jonathan Reproducibility of Aortic Pulse Wave Velocity Measurements Obtained with Phase Contrast Magnetic Resonance (PCMR) and Applanation Tonometry (P8) 96 80 56 33 79 46 122 82 108 126 T Tan, Ek Tsoon High Resolution Fast Inversion Recovery MRA (FIR-MRA) MR-Angioclub East Lansing 2009 74 153 Author Index V Varani, James Extracellular matrix metabolism in organ-cultured skin from patients with end-stage renal disease: Response to gadolinium based MRI contrast agents Velikina, Julia Design of Compressed Sensing Reconstruction for Highly Accelerated Time-Resolved MR Angiography Versluis, Bas MR Angiography of muscular and collateral arteries in peripheral and arterial disease: reproducibility of morphological and functional vascular status Voth, M Peripheral MRA with Continuous Table (CTM) Movements in Combination with High Temporal and Spatial Resolution TWIST – MRA with 0.1 mmol/lg Gadobutrol at 3.0 T 59 30 113 45 W Wang, Jinnan Improved Intraplaque Hemorrhage Detections by a Phase Sensitive IRTFE (SPI) Sequence Wang, Kang 3D Time-Resolved MR Angiography of Lower Extremities using Cartesian Interleaved Variable Density Sampling and HYPR Reconstruction Wang, Yi 3D peripheral vessel wall MRI with flow-insensitive blood suppression and isotropic resolution at 3 Tesla Magnetic source MRI for quantitative brain iron mapping Wieben, Oliver Comprehensive PC MR Imaging in Congenital Heart Disease Willinek, Winfried 4D-MRA in combination with arterial spin labeling for selective and functional information in patients with AVMs Wilson, Gregory Motion-compensated, flow-independent, non-contrast enhanced renal MR angiography (P9) Wright, Katherine Simultaneous Renal Angiography and Perfusion measurement Using Time-Resolved MRA Wu, Yijing Low Dose HYPR FLOW 37 48 41 109 88 70 127 100 31 X Xie, Jingsi Feasibility of Whole-Heart Coronary MRA on 3 Tesla Using Ultrashort-TR SSFP VIPR 91 Y Yuan, Chun Carotid Plaque Imaging and Clinical Risk Assessment 154 MR-Angioclub East Lansing 2009 34 Author Index Z Zhu, David The 3D SHINE Sequence Optimizes the Quantification of Carotid Intraplaque Hemorrhage Zwart, Nick 3D Dual VENC PCMRA using Spiral Projection Imaging MR-Angioclub East Lansing 2009 36 75 155