Download 01_Hampton Lectures 2011 Harris

3D Imaging Service
and Tumor Imaging Metrics Core
Gordon J. Harris, Ph.D.
Director, 3D Imaging Service and
Radiology Computer Aided Diagnostics Laboratory,
Massachusetts General Hospital
Director, Tumor Imaging Metrics Core
Dana-Farber/Harvard Cancer Center
Professor of Radiology,
Harvard Medical School
Hampton Symposium, October 15, 2011
Presentation Overview
•  Clinical 3D Imaging Service
•  Tele3D
•  3D Imaging Research and Development
•  Tumor Imaging Metrics Core
Gordon J. Harris, PhD, Director
Radiology Computer-Aided Diagnostics Laboratory
(RAD CADx LAB), 3D Imaging Research
Hiro Yoshida, PhD
Director, 3D Imaging Research
(Associate Professor)
Wenli Cai, PhD
R&D Faculty
(Assistant Professor)
Janne Nappi, PhD
R&D Faculty
(Instructor)
3D Imaging Service / Tele3D
Jennifer McGowan, MM, RT(R),(CT)
3D Operations Manager
Joyce Miller, AS
Administrator, Research
(Admin II)
Deirdre Pierce Tripoli, BS
Administrator, 3D/TIMC
(Admin III)
Tumor Imaging Metrics Core (TIMC)
Fresnel Josaphat, MS, Esq
Technical Director
Anthony Ranson, AS
Technical Support Specialist
Vadim Frenkel, MS
Web/Database Developer
Trinity Urban, BS
Operations Manager
Peggy MacWilliams
RT(R)(CT)
3D Technologist (Level II)
Joe Hakim, AAS,RT(R)(CT)
3D Technical Manager
Velda Singleton
25 New Chardon reception
Sally Pinho, RT(R)(CT)
3D Technologist (Level II)
3D Imaging Research
Yin Wu, Ph.D.
Principal Comp. Scientist
Julie Della Monica
RT(R)(CT)
3D Technologist (Level II)
Joan Beaudain, RT(R)(CT)
3D Technologist (Level II)
Diane Severini, BS, RDMS
3D/Clinical US Technologist
Hua Li, MD
3D Image Analyst
Shirley Thurston, BS
3D Image Analyst
Joan Davis, RDMS
3D US Technologist
Lea Hodge, RDMS
3D US Technologist
Jane Kelley, RT, RDMS
3D US Technologist
Adrian Ciobanu, RDMS
3D/Clinical US Technologist
Anand Kumar Singh, MD
Research Fellow
3D Imaging Service/Tele3D
KC Jackson, RT(R)(CT)
3D Technologist (Level II)
Vikram Kurra, MD
Research Fellow (BWH)
Gisele Cruz, MD
Research Fellow (MGH)
Marta Braschi, MD
Research Fellow (MGH)
TBN
TIMC Coordinator
George Oliviera M.D.
Per diem
Tumor Imaging Metrics Core
Cheryl Sadow MD
Quality Officer (BWH)
Koichi Nagata, MD, PhD
Research Fellow
June-Goo Lee
Research Assistant
William Hanlon, MS
Project Manager
Services
•  3D Imaging Service
– MGH Radiology, 12 years, started February 1, 1999
•  Tele3D (New External Service)
– National/International 3D Imaging Service, started 2009
•  Tumor Imaging Metrics Core
– Dana-Farber/Harvard Cancer Center (5 Harvard
Hospitals), initiated 2004
•  Neurofibromatosis Imaging Analysis Archive
– National: MGH, JHU, UAB, CNMC, OSUMC, HEI, CTF
Grants
•  Alcoholism Brain Imaging Research, NIH
– Collaboration with BUSM, CNY, in 9th year, renewed ->
2014
•  SPOTRIAS, NIH Program Project, Neuroimaging Core
•  Tumor Imaging Metrics Core (DF/HCC)
•  Neurofibromatosis Imaging Grant, PET, MR Tumor
Volume
– MGH, Germany collaboration, DoD
•  Harvard Catalyst, CTSC, Imaging Subcommittee
•  CAD grants, CT Colonography (Hiro Yoshida and
team)
Goals of 3D Clinical Service
•  Integrate Computer Aided Diagnosis
(visualization, quantitative analysis) into
routine clinical workflow
•  Bridge research and clinical applications
to migrate new technologies into clinical
practice
3D Imaging
Axial images are ‘stacked’ into a threedimensional volume
Stacked Axial Slices
sjn/MGH
3D Volume
Information Transmission
Courtesy of Matthew A. Barish, M.D.
Information Transmission
Courtesy of Matthew A. Barish, M.D.
CSF Leak
Image Processing Techniques
•  Multi Planar Reconstruction (MPR / MPVR)
•  Oblique and Curved Reformat
•  MIP (Maximum Intensity Projection)
•  Shaded Surface Display
•  Volume Rendering
•  Endoluminal Views (Virtual Colon, Bronch, Vessels)
•  Functional Imaging
•  CADx: Segmentation, Quantitation
MIP Rendering
•  Maximum Intensity Projection
•  Displays the pixels of highest intensity along a ray
•  Useful for vascular anatomy
sjn/MGH
Orthopaedics:
Volume Rendering (VR)
Vascular Imaging:
VR
Curved
CT Bronchography:
Endoluminal
See-Through
Neurovascular CTA/MRA:
VR
MIP
Cardiac CTA:
Normal
3˚CABG
Functional
CT Perfusion
CBV
CBF
MTT
Functional MRI (fMRI)
Computer-Aided Segmentation
Virtual Hepatectomy
Semiautomated segmentation
Total Volume = 1551.8cc’s
Virtual Hepatectomy
Transplant segment
•  Define surgical plane
•  Sufficient volume for
regeneration
Left Lobe segment 316.91 cc’s
CAD: Automated Segmentation:
VR
MIP
Clinical Benefits of 3D Radiology
•  More comprehensive and realistic views
of patient
•  Faster, more confident diagnoses and
treatment planning decisions
•  Reduced need for exploratory surgery
•  Minimize surgical invasiveness and
operating room time; reduced damage to
healthy tissue
3D Imaging Service at MGH
•  Clinical computer visualization for
routine clinical use; 3D on request.
•  Fast turnaround, full-time
technologists
•  Full integration with hospital PACS
and information, billing systems
•  Currently, over 2000 cases
processed per month
MGH 3D IMAGING SERVICE:
DAILY AVERAGE VOLUME
BY YEAR
140.0
120.0
100.0
80.0
Series1
60.0
40.0
20.0
0.0
FY' 99
FY' 00
FY'01
FY'02
FY'03
FY'04
FY'05
FY'06
FY'07
FY'08
FY'09
FY'10
3D Imaging Service Hardware
and Software at MGH
 
5 GE Advantage Workstations
 
5 Vital Images Vitrea Workstations
 
Tera Recon Aquarius Workstation and NetServer
 
2 Voxar Workstations
 
2 LINUX Computers with MedX/VolumePro for fMRI
 
Mirada Fusion7D Workstation for image registration
 
3 GE LogiqWorks for 3D Reconstruction of US
 
AGFA PACS Service Station, and RadWorks PACS
 
Materialize SimPlant/Master for Dental Implant Planning
 
MMS Preview LINUX Server
 
LINUX DICOM Server, 7 PCs, 2 Macs
Tele3D Service
•  Many facilities do not possess the
resources or expertise to manage or
maintain a 3D Imaging Service
•  At low volume, it is not cost-effective to
maintain the staff, training, infrastructure
•  We developed the Tele3D Service to
provide access to the resources of the
MGH 3D Imaging Service to outside
hospitals and imaging centers
TELE3DADVANTAGE
Setting a Higher Standard in 3D  2. Clinical Benefits: Workflow and Productivity
Typical 3D Workflow
Scan
Patient
QA Exam
Store on
RIS / PACS
File
Report
Read
Exam
3D Workstation
30
TELE3DADVANTAGE
Setting a Higher Standard in 3D  2. Clinical Benefits: Workflow and Productivity
Tele3D Workflow
Referring Physician
Client Radiology Department
Final Report
for Review
Exam to
PACS
Order Study
with 3D
Final Report
to RIS
2D & 3D Images
RIS/
PACS
for Interpretation
Tele3D 3D
Service
3D Images
Order 3D Exam
Gateway
2D Images
31
TELE3DADVANTAGE
Setting a Higher Standard in 3D  2. Clinical Benefits: Workflow and Productivity
Tele3D’s Workflow Benefits
Radiologists
•  Radiologists are able to focus on
patient care, teaching, and
research rather than generating 3D
images
Technologists
•  Technologists are free to focus on
efficiently scanning patients to
increase throughput
32
TELE3DADVANTAGE
Setting a Higher Standard in 3D  1. Tele3D Overview
Complete Range of 3D Protocols
Vascular Radiology
Neuroradiology
Cardiac Radiology
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
CTA/MRA for AAA
CTA / MRA for TAA
CTA / MRA for aortic dissection
Abdominal MRA for mesenteric
ischemia
Abdominal / pelvis MRA / MRV
for portal & deep vein
thrombosis
MRI Upper Extremity Runoff
CTA Runoff
MRA Chest
MRI Vascular Run-off
Renal MRA for Stenosis
Bone & Joint Radiology
• 
Skeletal fractures (Spine, Face,
Temporal and Joints)
• 
Head CTA / MRA
Neck CTA / MRA
Head CT / MR Venography
Head CT / MR Perfusion
Facial Bones
Mandible CT for inferior
alveolar nerve
Pediatric CT for
craniosynostosis
• 
• 
• 
Cardiac CTA
Chest CT for pulmonary vein
evaluation prior to ablation
Chest CT for Pulmonary
Arteries
Chest MR for pulmonary vein
evaluation prior to ablation
Cardiac calcium scoring
Abdominal Radiology
Chest Radiology
• 
• 
Chest CT for any tracheal
lesion (Virtual Broncoscopy)
Chest CT for Video Assisted
Thoracotomy (VAT) planning
• 
• 
• 
• 
• 
• 
• 
Liver Resection / Liver Donor
CT Liver Volumes
CT Urography / Hematuria
CT Renal Donor
MRCP
Pancreas CTA
MRI Liver and Spleen
volumes for Gauchers Disease
33
3D Imaging Research and
Computer Aided Detection (CAD)
•  MGH 3D Imaging Research and CAD
team includes 5 Faculty PhD/MD staff plus
fellows
•  Funding primarily through research
grants
•  Focus on developing clinical imageprocessing applications
34
Liver Segmentation (transplant)
CAD Lab
Liver-Tumor Segmentation
CAD Lab
0 months
2.3 months
5.7 months
13.8 months
18.6 months
19.6 months
In-Vivo Segmentation
•  In-vivo segmentation by CAV (DT level set)
CAD Lab
Renal Stone Segmentation
CAD Lab
NF-Tumor Burden
Tumor Burdon
2
4
80
17
336
12
3
257
570
9
6
10
Number of
Tumors
(Total: 63)
CAD Lab
155
73
772
252
Volume of
Tumors (CC)
(Total: 2495)
Automated Pneumothorax
Quantification
741.18 CC
63.50 CC
An example of the CAV scheme for the quantification of
pneumothorax on a 12-year-old/male patient with
metastatic angiosarcomas, who had pneumothorax after
pleural biopsy.
CAD Lab
Sarcoma Viable Tissue Quantification
T2 Fat Sat Sequence
15 Sec
CAD Lab
25 Sec
70 Sec
Early Enhancing
180 Sec
Meniscus Tear
CAD Lab
Meniscus
Superior View
CAD Lab
Inferior View
Electronic Cleansing (EC)
•  “Virtually” cleanse the colon in CTC images
Before EC
CAD Lab
After EC
Mosaic Decomposition Example
CAD Lab
Fecal-tagging image
Tile decomposition
Tile classification
Cleansed image
Electronic Cleansing
CAD Lab
Detection of Polyp Candidates
polyp (cap)
colonic
wall
CAD Lab
shape index
fold (ridge)
Tumor Imaging Metrics Core
A Centralized Service to Provide Standardized Tumor
Measurements for Oncology Clinical Trials
www.tumormetrics.org
25 New Chardon St., Suite 400C,
Boston, MA 02114
Why TIMC?
Current Issues
TIMC Solution
• 
Clinical read ≠ Response criteria
• 
Standardized measurements
• 
Inconsistent measurements due to
different readers across time points
• 
Centralized image analysis to
increase reliability
• 
Time-consuming for trial staff &
radiologists
• 
Online data entry & review system to
increase efficiency
• 
No image-based longitudinal record
• 
Web-based secure results reporting
• 
Image data not available for audit
• 
Centralized database and archive
49
TIMC Experience
•  Service BWH, DFCI, MGH,
BIDMC and CHB
Assessment Criteria Breakdown
•  >300 active trials
RECIST 1.0
72%
RECIST 1.1
7%
•  >17,000 scans analyzed
Cheson/IWRC
10%
PET SUV
6%
WHO
3%
Choi
1%
Brain Volumetrics
1%
•  Proficient in various
assessment criteria
•  Most new trials now budget
for TIMC services
50
TIMC Exam Volume
*projected
51
Usage by Disease Group
52
Usage by Institution
53
TIMC Services
Protocol Consultation
•  Consultative services for imaging components in trials
Image Capture & Analysis
•  DICOM transfer from DF/HCC institutions’ PACS
•  Import of outside scans via secure internet or CD
•  MRI, CT, PET & multi-modality scans
•  Standardized linear, volumetric, SUV and density
measurements
Web-based Services
•  Online Order-Entry for authorized users
•  Results viewable by study staff on a secure website
•  Centralized database for internal/external auditing
54
TIMC Staff
Co-Director:
Co-Director:
Director of Quality Control:
Project Manager:
Operations/Projects Manager:
Research Coordinator:
Administrative Assistant:
Research Fellow:
Research Fellow:
Research Fellow:
Research Instructor:
DBA / Web Developer:
Technical Director:
Technical Support:
MGH
Gordon J. Harris, Ph.D.
Annick D. Van den Abbeele, M.D.DFCI
BWH
Cheryl Sadow, M.D.
MGH
William B. Hanlon
MGH
Trinity A. Urban
MGH
TBN
Deirdre Tripoli MGH
Marta Braschi MGH
Gisele Cruz, M.D. BWH
Vikram Kurra, M.D. BWH
Anand Singh, M.D.MGH
MGH
Vadim Frenkel
MGH
Fresnel Josaphat
MGH
Anthony Ranson
55
Contact Information
For more information, please visit our
website at: www.tumormetrics.org or
email us at: [email protected]
56
Impact of New Technologies
on Radiology
 
 
 
3D Visualization enables faster, more
confident diagnoses and treatment
decisions
Quantitative analysis and CAD can
provide more accurate, reliable treatment
planning, staging, and assessment
Improved patient care, increased clinical
confidence, reduced time, cost, and
invasiveness
Any
Thank
Questions?
you!!!