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
Robert Jonathan Jeffrey R. Edelman, Kleefield, B. Mendel, MD MD MD #{149} Heinrich #{149} Craig #{149} Dennis Extracranial with “Black Carotid Blood” The authors evaluated the accuracy of “black blood” magnetic resonance (MR) angiography for depicting disease involving al carotid the arteries. Two- extracrani- and three- dimensional flow-compensated gradient-echo sequences were ployed to create “bright blood” ages. A thin-section quence with flow lowed the creation images. made images mumjection angiograms bright with emim- spin-echo sepresaturation alof black blood Projection from and were black application blood of a maxi- or minimum-intensity algorithm, pro- tients, both carotid arteries with both bright methods; in pawere sensitive in 13 of 33 arteries, mostly in severe disease; this problem was not encountered with black blood angiography. The authors conclude that bright blood angiography is a sensi- tive method for screening carotid disease; when a significant abnormality is found, black blood angiography should be performed for more precise delineation of the lesion. Index terms: Carotid arteries, MR studies, 172.i214 #{149} Carotid arteries, stenosis or occlusion, 172.721 #{149} Magnetic resonance (MR), image processing, 172.1214 #{149} Magnetic resonance (MR), pulse sequences, 172.1214 Radiology I From 1990; the the disease cause carotid arteries of stroke. Several affecting is a major imaging mo- dalities have been used for diagnosing carotid artery disease. Conventional angiography is considered the standard of reference but is invasive. #{149} its accuracy of the carotid SUBJECTS AND for bifur- METHODS performed imaging at 1.5 T system (Mag- vasive method for screening suspected carotid disease but has a limited field of view (FOV), is degraded by calcified plaque, and may not enable the distinction of severe stenoses netom SP; Siemens un, NJ). A Helmholtz Medical coil, Systems, Isewhich con- was the studied. transmitter. Thirty subjects These included 13 healthy were vol- from unteers (mean 23-48 sonography is a good nonin- occlusions. techniques creating have magnetic (MR) angiograms in a projectional that deformat studies, “bright blood” imag- ing methods have been used (ie, flowing protons are made to appear more intense than surrounding tissues). However, there are substantial technical problems with bright blood methods. These problems include the exaggeration recirculation of stenoses due and turbulence the stenosis. In addition, stenoses may cause low ties and therefore We report an for evaluating very severe flow veloci- poor alternative the to flow distal to flow carotid contrast. approach arteries with conventional years) without tery disease. conventional for sequence 17 patients angiog- B.W., RB., J.K., J.B.M., D.J.A.) and age, 35 years; any history years) were went selective did 58 years; who rial inch respectively). digital image within with- the Committee (9X in all cases Studies were guidelines on and or intraarte- angiography with a 1,024 1,024 matrix) was performed in at least two projections. hospital range, under- (14 patients) (three patients Cut-film subtraction intensifier performed ar- undergo angiography in 1 day of the MR study within 2, 6, and 17 days each, not and was used In addition, studied carotid range, of carotid group angiography development. (mean age, 24-77 Clinical of the Investiga- tions. Bright Blood Angiography In all subjects, bright blood angiograms were obtained before black blood angio. grams. Earlier studies sagittal employed a single- three-dimensional gradient-echo tients). Imaging (GRE) sequence variables were tion of 40 msec, time (TR) of 8 msec (TR/TE 15#{176}, and one 70/8, 25#{176} flip thickness, 256 two X 192 patient 40/8), excitation angle, for a two-slab slab thickness, tion H.P.M., of the neck, was used and the body coil This or two-slab with “black blood’ angiography. With black blood angiography, flowing protons are made to appear hypointense, to reduce artifacts created by abnormal flow patterns. In this study, black blood angiography was compared forms to the shape for signal detection, In (R.R.E., MD raphy to determine depicting disease cation. tion, of Radiology Bajakian, All studies were with a whole-body Ouplex 177:45-50 Departments #{149} Richard MD Arteries: Evaluation MR Angiography’ THEROSCLEROTIC these for detecting carotid disease. However, bright blood angiography exaggerated the severity of carotid lesions MD Wallner, Skillman, similar to that of conventional angiograms. MR angiography has been used with some success in the evaluation of carotid artery disease (1-8). In and conwas per- methods #{149} Bernd J. MS resonance pict vessels These methods were used in 13 healthy volunteers and 17 patients, and a prospective blinded compari- formed. Normal were well shown and black blood MD #{149} John Recently, several been described for respectively. son of MR angiography ventional angiography P. Mattle, Kent, MD J. Atkinson, first-order studies, echo time (TE) flip angle of for a single slab and one excitation acquisition, 32 partitions, matrix, (six paa repeti- with a 40-mm 1.25-mm secflow and or compensa- a 23-cm FOV. a three-dimension- Surgery (C.K., J.J.S.), Beth Israel Hospital and Harvard Medical School, 330 Brookline Ave. Boston, MA 02115; Department of Radiology, New England Deaconess Hospital, Boston (H.P.M.); and Siemens Medical Systems, Iselin, NJ (D.J.A.). Received March 15, 1990; revision requested April 19; revision received May 30; accepted May 31. Address reprint requests to R.R.E. c) RSNA, 1990 Abbreviations: FOV field of view, GRE gradient echo, RF = radio frequency, SE spin echo, TE = echo time, TR repetition time. 45 al sequence with an axial excitation and coronal readout was tested. This technique excites through only the a limited neck, axial without volume affecting the great vessels within the chest. As a result, flow contrast is improved. Imaging vanables were 40/9, 20#{176} flip angle, one excitation, 10-cm axial excitation volume, double presaturation slabs positioned to eliminate wraparound artifact from the anterior and posterior neck and to reduce jugular venous signal, first-order flow compensation, 256 X 192 matrix, 23-cm FOV, 64 partitions yielding a 1.5-mm section thickness, and an 8.2-minute imaging time for both carotid arteries. In 11 patients, sequential multisection two-dimensional acquisitions were performed. A series of sagittal multisection two-dimensional flow-compensated GRE images were obtained, with three sections per ac- quisition. Imaging variables were: 72/10, 350 flip angle, two excitations, 3-mm section thickness with a 6-mm intersection gap, and a 256 X 256 acquisition matrix with a 23-cm FOV. Two patients under- a. went Figure bright both two- blood and three-dimensional angiography. nies Bright blood images were cessed with a maximum-intensity tion algorithm. A series ages were generated individual images, for evaluation. Black Blood Several postpro- of projection of MR imthe open angiograms sequential of healthy multisection two excitations, the skull. The atlas loop. (b) gonithm. Long projec- and, along with were photographed b. 1. curved subjects. two-dimensional (a) Bright bright blood blood angiogram images created (72/ 10, 3-mm sections). The left carotid artery is seen from its origin left vertebral artery is also seen from its origin at the subclavian Black blood angiogram postprocessed with a minimum-intensity straight arrow internal carotid artery, short arrow external arrow = jugular vein, solid curved arrow = vertebral artery. from 35#{176} flip a se- angle, to the base of artery to the projection alcarotid artery, Angiography methods were studied in healthy subjects, to create black blood images of the carotid bifurcation. The requirements for these sequences were that (a) they produce a uniform signal-intensity void within the carotid artery, (b) good contrast be maintained between the yessel and surrounding fat and muscle, (c) susceptibility and chemical shift antifacts be minimal so as not to interfere with plaque visualization, and (d) imaging times be sufficiently short that gross patient motion is unlikely to occur during the study. The two- and three-dimensional GRE sequences we tested with flow presaturation (9,10) failed to satisfy requirements and therefore are not scnibed in detail here. The most effective method tested two-dimensional with flow spin-echo presaturation (SE) and thin these detations, was sequence sections. The use of a thin section exacerbates signal loss from washout, which occurs as flowing protons move out of the section during the time interval TE/2 between the 90#{176} and 180#{176} radio-frequency (RF) pulses. Initial sequence comparisons included variations in section thickness (1.2-2.0 mm), intersection gap (0%-40% of section thickness), and sampling times (7.7 vs 12.8 msec). The resulting images were judged according to the four criteria listed previously. On the basis of these initial sequence comparisons, imaging variables considered to represent the “optimal” black blood technique were a doubleecho SE sequence (1,600/20, 45), two exci- 46 #{149} Radiology a 7.68-msec sampling 192 acquisition matrix, msec timized with use profiles, RF pulses section thickness sagittal with acquisition no time, 23-cm 256 FOV, X 6- of computer-op1.8-mm section intersection with two gap, groups slab tion, Flow was presaturation employed applied just presaturation and of 2 mT/rn, and of 8-mT/m tion. The encompassing with a constant amplitude total spoiler and tion algorithm of 5-msec implemented on (Digital A series ages were is projec- the most hypenin- a user-defined viewing into the projection im- of black blood generated with projection viewing of -30#{176}to +30#{176} and were along with the individual imangles photographed images for evaluation. dura- imaging time of a study both carotid bifurcations II computer than pixel along is extracted age. technique that has been previously except that the most hy- rather tense, angle This maximum-intensity gradient was 10.2 minutes. The black blood images were cessed with a minimum-intensity MicroVAX Mass). to the pointense, of the carotid a 5-cm axial caudad to the bifurcaselection gradient Maynard, tion algorithm described (1 i), sections acquired simultaneously and each group centered over a carotid bifurcation. arteries ics, similar Data Analysis The studies postproprojec- were graded mild stenosis the tery the host Electron- (less sagittal of the carotid as follows: of the than 50% plane), internal bifurcation 1 normal, 2 carotid an- diameter reduction 3 moderate stenosis October in 1990 Figure 2. blood Comparison of bright angiography. (a) gram shows a severe stenosis the left internal carotid artery. blood angiogram created tial two-dimensional sion at the origin Black blood black angio- at the origin (b) Bright with the approach of the internal tery but shows a finding of a second stenosis more (c) and Conventional sequen- shows a lecarotid ar- falsely distally angiogram of suggestive (arrow). more correctly depicts a solitary stenosis (arrow). (d) Conventional angiogram in the same patient shows a mild stenosis of the right internal carotid artery. (e) Bright blood angiogram accurately depicts the mild stenosis (arrow). (f) Black blood angiogram also correctly depicts the stenosis (arrow). a. b. C. giograms the ses. in these length In one proximal cases exaggerated and severity of the stenosevere stenosis of the internal carotid artery (Fig 2b), findings at bright blood angiography suggested a second stenosis more distal; this was not shown at black blood or conventional angiography (Fig 2a, 2c). Generally, there was close correspondence between the appearance black d. of the internal carotid diameter reduction in 4 = severe stenosis of artery (more than 75% artery (50%-75% the sagittal plane), the internal diameter carotid reduction in the sagittal plane), and 5 = occlusion of the internal carotid artery. Two authors (R.R.E., H.P.M.) prospectively interpreted the MR studies without knowing the results of conventional angiography. Both the bright and black blood studies were available for this in- terpretation; the projection angiograms and the individual sections were exammed. Two other authors (J.K., J.B.M.) reviewed the conventional angiograms without knowing the results of MR angiography. In each case, the interpretation was made by consensus servers. of the two ob- In one case, ancy had been found ventional angiogram rotid occlusion) (which showed giography was after a major discrepbetween the con(which showed a caand the MR angiograms a severe stenosis), MR an- repeated. RESULTS For black blood imaging, quences were ineffective flow voids, despite uration. Lengthening crease flow-related ed in a loss of vessel sel-muscle contrast. SE sequences proved Volume 177 #{149} Number the GRE sefor creating use of presatthe TE to indephasing resultdetail and yesTwo-dimensional effective for 1 creating The black flow blood void images was (Fig maximized 1). with the use of very thin sections and no intersection gap. Because we desired to maintain imaging times of 10 mmutes or less and images became noisier as the section thickness was reduced, a section thickness of 1.8 mm was found to represent promise between volume averaging a good minimizing and compartial an adequate signal-to-noise ratio. The sequence with a shorter sampling time was preferred because chemical shift artifact in images obtained with the lower-bandwidth sequence could be falsely interpreted as plaque. The results of the clinical study are summarized found was in the Table. at conventional detected with All disease angiography MR angiography. In 20 arteries, the bright blood and black blood angiograms helped correctly grade and provide similar portrayals of normal carotid bifurcations and of mild to moderate stenoses (Figs 1, 2). Bright blood angiograms helped detect all stenoses of the internal carotid artery. However, in 13 arteries, only grams usually correctly in severe (Figs 2-4). sions and grams. In two patients, giograms failed signal intensity f. e. of the blood the black blood angio- depicted the disease, stenoses or occluThe bright blood an- carotid artery occlusion, conventional stenosis on conventional the angio- bright blood anto demonstrate high within the internal and falsely whereas black angiograms strated a severe images of another stenosis. cluded internal suggested blood and demon- Bright patient carotid blood with an oc- artery showed mildly high signal intensity within the vessel that could be incorrectly interpreted as slow flow. Black blood images showed uniform intermediate signal intensity within the vessel, a finding consistent with thrombus. In one patient in whom both bright blood and black blood angiography showed severe bilateral stenoses of the proximal internal carotid conventional angiography, performed 4 days later, showed a severe stenosis of the right internal carotid artery and an occlusion of the left internal carotid artery (Fig 4). Because of the discrepancy, a repeat MR angiogram was obtained 2 days later. Both bright and black blood angiog- artery, raphy showed again a severe stenosis of the right internal carotid artery but an occlusion of the left internal carotid artery. Black blood images showed thrombus with a peripheral rim of sity filling ternal high moderately carotid the signal lumen of artery (Fig the intenleft in- 4e). Radiology #{149} 47 DISCUSSION able Bright tool blood angiography for the depiction is a valuof vascular anatomy and function. The technique has proved advantageous for the evaluation of disease involving the intracranial vessels, such as arteriovenous malformations (12,13), eurysms (14), and venous disorders (15). Recent work has suggested ana po- tential role for MR angiography in the assessment of carotid bifurcation disease (16,17). However, the anatomic bright depiction blood cise, of vascular angiography especially due in severe to signal lence, loss recirculation, raphy flow or (in the study, trast An case bright blood between the perivascular that for f. de- but of was con- vessel by e. angiog- approach creates blood tissue c. of in all cases the degree alternative developed b. slow to be sensitive picting carotid disease frequently exaggerated disease. therefore a. turbu- acquisitions) proved at stenoses, from three-dimensional flow. In our lesions is impre- and rendering flowing blood as a signal-intensity void. We designated this technique as black blood blood angiography. angiography in that the flow signal produced does not In our study, nique ages for producing combined flow with an SE sequence sections. For the section, ity that will the flow the tech- black blood presaturation im- and very thin perpendicular maximum result ensure to flow in an thickness)/(TE/2). a section thickness TE of 20 msec, the will the is flow best intensity on an by the ratio alone by degrade Also, good flow contrast even at relatively low velocities. signal termined an advantage loss turbulence images. preserved Black has veloc- observable is de- Vmax (section For instance, with of 1.8 mm and a washout effect flow velocity to 8 cm/sec, 18 the most pected to occur an impractically al sections compass fore, a sagittal be carotid #{149} Radiology needed artery. orientation ployed (19). Because of the 48 is ex- for an axial section, large number of would the washout very thin its origin. (c) stenosis of Black (arrow) the blood at right the carotid angiogram shows origin of the artery shows occlusion external carotid severe stenosis of the internal artery. (d) (approximately carotid Digital artery subtraction 75% reduction in diameter) at the origin of the internal carotid artery and occlusion of the external carotid artery. (e) Sequential two-dimensional bright blood angiogram shows reconstitution of flow signal intensity within the internal carotid artery (arrow), consistent with turbulent flow distal to a stenosis. The stenosis, however, is poorly depicted. (f) Black blood angiogram shows a severe stenosis at the origin of the internal carotid artery and occlusion of the external carotid artery. The stenosis appears slightly less severe in the individual black blood im(not shown). exaggerate the Both extent the maximum- of vascular and minimum-intensity projection algorithms tend to disease. sections, flowing protons are still likely to wash out of the section between the 900 and i800 RF pulses. Signal loss due to flow turbulence generally enhances, rather than degrades, the quality of black blood images, which is particularly relevant in axi- to en- Therewas at the al- though the signal-to-noise ratio is suboptimal in the second-echo images. In comparison, the mean flow velocity measured in the internal carotid artery with Ooppler sonography is 36.3 cm/sec ± 8.6 cm/sec (18). Although seen and ages void if the velocity is greater than cm/sec. The availability of a second echo (TE = 45 msec) reduces the minimum Studies in a patient with an occlusion of the left internal carotid artery and moderate stenosis of the right internal carotid artery. (a) Conventional angiogram of the left Carotid artery shows occlusion of the internal carotid artery and stenosis (arrow) at the origin of the external carotid artery. (b) Sequential two-dimensional bright blood angiogram demonstrates occlusion of the internal carotid artery, but the external carotid artery is poorly 3. angiogram SE image a complete d. Figure em- (1.8 mm) the These black better in evaluation of severe stenoses. factors may explain why the blood techniques performed than bright blood techniques depicting severely Several drawbacks blood technique are stenotic vessels. of the black also evident. the cannot be applied the carotid artery that pass through bone (eg, the petrous portion), because of the absence of signal adjacent to the artery. Second, First, to portions method of the useful field of view is limited; the bright blood images were much better for studying the common carotid artery. Third, in some projections, overlap of the carotid ular vein could limit problem is prevented the individual black well as the projection and jug- evaluation. by inspecting blood images angiograms; artery This October as 1990 a. b. C. Figure 4. Studies of a patient with duplex of a severe stenosis artery. (a) Sequen- sonographic evidence the left internal carotid tial two-dimensional bright blood of angio- gram origin depicts severe stenosis (arrow) at the of the internal carotid artery. (b) Black blood angiogram also shows the severe stenosis (arrow), which appears shorter than in a. (c) Conventional angiogram obtained occlusion 4 days later left internal of the (d) Repeat bright blood shows interval carotid artery. angiogram, obtained 2 days after the conventional angiogram, shows the occlusion. (e) Single section from repeat black blood acquisition also shows the occlusion and high signal intensity (arrow) around the acute thrombus. slow flow in the case of critically stenosed vessels. Compared with threedimensional methods, the advantages of sequential two-dimensional methods d. C. multiplanar reconstructions be helpful. When in the either. projection Vessel prevented presaturating tense with images. plaques is more is the possibility plaque could blood seen flowing in our blood study intensity from flowing within the vessel lumen. A limitation of our study - 1’7’7 #{149} images by vein. Finalthat densebe isoin- Nonetheless, termediate signal readily distinguished er-signal-intensity blood imaging, quential sagittal also to ex- angiograms overlap in bright blood the jugular ly, there ly calcified easily is taken dude the most lateral sections taming the jugular vein from processing, overlap is seldom lem conposta prob- was bright the mal has can care I that the methods employed for blood imaging evolved over course of the study, and an optibright blood imaging sequence not yet been devised. For bright because imaging and it was sensitive times and ately rotid specific disease. most advantage compared all dimensional showed and were the lowblood in- for depicting The major with sequential imaging were moder- axial (17) is seap- catwo- a shorter might However, be better axial for acquisitions detecting very better flow contrast in the setting [20]) and motion for sections. Three-dimensional the dual advantages and smaller voxels of low less sensitivthe individ- ual ner sections), from Masaryk loss sults with so sequences of shorter (due to the there is less have TEs thinsignal flow-related et al (16) dephasing. obtained good a sagittal three-dimension- al sequence in study time, since much fewer sections are required. The sagittal orientation uniformly provided good flow contrast when thin sections were ob- tamed. design the proach short in black the we preferred two-dimensional are (particularly flow velocities ity to patient by imaging their a 30-cm-diameter, ceive-mode, re- patients transmit-re- linearly polarized head coil. Since inflowing spins beyond the limit of the head coil are not affected by the RF pulses, flow contrast is generally adequate. ence with rotid bifurcation disappointing, the Our linear head imaging because experi- coil for cahas been the bifurcaD.%A;,.i,%,._ AG tion is not consistently within the imaging coil in large short fore, coil patients necks. id artery or those Also, the is poorly we used with positioned volume of the common the body coil carot- receiver and coronal read-out. However, the resulting coronal ages were suboptimal ing carotid stenoses; tended to be degraded We currently the sequential sagittal A final point is that created projection tracranial Therefore, im- supplement vidual images most precise angiograms an overall particularly projection noses bright blood angiography depiction and plaque Bright Wedeen RA, Edelman imaging of pulsatile magnetic resonance. Science RR, flow Dumoulin CL, phy. 6. 7. E. Magn Souza SP, Walker Time-resolved Reson Med 14. 9. 17. DG, 18. Yo- JM. by lective inversion recovery. Magn Reson Med 1988; 7:472-484. Edelman RR, Atkinson DJ, Silver MS. FRODO pulses: a new method for elimiof motion, Radiology flow, 1988; and se- with selective venography. 19. clinical Dumoulin with and in progress. 6:275-286, Macovski A, Pauly of MR angiography and gy 1989; 171:801-806. Keller PJ, Drayer BP, ography flow im- MR arteri- Radiology 1989; experience. Fram CL, Souza EK, SP. two-dimensional Williams MR angiacquisi- three-dimensional Radiology Radiolo- display: 1989; work 173:527-532. Mattle H, Grolimund P. Huber P. Sturzenegger M, Zurbrugg HR. Transcranial Doppler sonographic findings in middle cerebral artery disease. Arch Neurol 1988; 45:289-295. Daniels DL, Kneeland JB, Foley WD, et al. Cardiac-gated local coil MR imaging of the carotid neck bifurcation. AJNR 1986; 7: 1036-1037. 20. 21. Nishimura Considerations 12:377-382. KR. MR Masaryk TJ, Modic MT. Ross JS, Ruggieri P. VanDyke C, Tkach J. MR angiography of intracranial aneurysms (abstr). AJNR 1989; 10:893-894. Mattle HP, Edelman RR, Reis MA, et al. Cerebral venography and flow quantification with MR. Radiology 1989; l73(P):187. Masaryk TJ, Modic MT. Ruggieri PM, et al. Three-dimensional (volume) gradientecho imaging of the carotid bifurcation: with 3:454-462. 8. evaluation preliminary MR angiogra1988; arti- 173:831-837. KD, Dumoulin CL, Souza SP, Walker MF, Wagle W. Three-dimensional phase contrast angiography. Magn Reson Med 1989; 9:139-149. Dixon WT, Du LN, Faul DD, et al. Projection angiograms of blood labeled by adiabatic fast passage. Magn Reson Med 1986; presaturaflow aging in vascular malformations using gradient recalled acquisitions. AJNR 1988; 9:637-642. Edelman RR, Wentz KU, Mattle H, et al. Intracerebral arteriovenous malforma- et al. MF, Spatial suppressing put Assist Tomogr 1988; Needell WM. Maravilla tion 5. RL. for facts and improving depiction of vascular anatomy in MR imaging. Radiology 1987; 164:559-564. Laub GA, Kaiser WA. MR angiography with gradient motion rephasing. J Com- ography 1985; 4. artifact. #{149} Radiology Meuli Projective nation 50 VJ, JP, Ehman a method tions: 16. 172:351-357. 230:946-948. Dumoulin CL, Hart HR. MR angiography. Radiology 1986; 161:717-720. as anatomy, angiography, black provides more acof severe carotid ste- morphology. 13. 15. Ruggieri PM, Laub GA, Masaryk TJ, Modic MT. Intracranial circulation: pulse-sequence considerations in three-dimensional (volume) MR angiography. Radiology 1989; 171:785-791. shitome In conclusion, bright blood angiography is a good screening method and accurately portrays normal bifurcations and mild stenoses. Compared with blood curate 12. U 2. 3. the projection useful to vascular the clinician. arteries. Edelman RR, Wentz KU, Mattle H, et al. Projection arteriography and venography: initial clinical results with MR. Radiology a of mdifor 1 1. of 1. an- with the maximumalgorithm show is required guide for carotid 1989; results. The are nonetheless or combination References an axial threethis combina- inspection of moderate The We are indebted to Gerhardt Laub, PhD, of Siemens Medical Systems, Erlangen, Federal Republic of Germany, for assistance in pulse sequence development, and Mark Haacke, PhD, of Case Western Reserve University. Cleveland, for helpful discussions. loss of vascular detail due to random noise propagation (21). We have found this to be true of the minimum-intensity projection algorithm as well. when angio- Acknowledgments: axial two-dimen- sional sequence with dimensional sequence; tion appears promising. disease. Felmlee tion: latter be employed bright blood are suggestive 10. are the these methods provides an accurate depiction of disease affecting the ex- for interpretalso, the studies if the patient swallowed. angiography methods; should on the severe as transmitter. excitation giograms intensity gram There- tried a three-dimenwith use of an blood method findings the saturation effects with three-dimensional sequences, we sional sequence black complementary with visualized. a Helmholtz To overcome encountered and Haacke EM, Masaryk TJ. The salient features of MR angiography (editorial). Radiology 1989; 173:611-612. Anderson CM, Saloner D, Tsuruda JS, Shapeero LG, Lee RE. Artifacts in maximumintensity-projection display of MR angiograms. AJR 1990; 154:623-629. wraparound 166:231-236. October 1990