Nuts and Bolts of MRA

Nuts and Bolts of MRA

3. Ernst EEW, Matria A. Exercise for intermittent claudication. Cardiol Board Rev 1988;5:82. 4. Lindgarde F, Jelnes R, Bjorkman H, et al. Conserva~ ti...

4MB Sizes 0 Downloads 16 Views

3. Ernst EEW, Matria A. Exercise for intermittent claudication. Cardiol Board Rev 1988;5:82. 4. Lindgarde F, Jelnes R, Bjorkman H, et al. Conserva~ tive drug treatment in patients willt moderately severe chronic occlusive peripheral arterial disease. Circulation 1989;80,1549-1556. 5. McDaniel MD, Cronewett J1. Basic data related to natural histolY of intermittent claudication. Ann Vasc Surg 1989;327')-277. 6. Sise MJ, Shackford SR, Rowley WR, et al. Claudication in young adults: a frequently delayed diagnosis. J Vase Surg 1989;10,68-74. 7. Nilson SE, Wolf GL, Cross AP. Percutaneous [ranslu~ minal angioplasty versus operation for peripheral arteriosclerosis. J Vasc Surg 1989;9:1-9. 8. Glover J1. Presidemial address: vascular surgerythe third generation. J Vase Surg 1990;1L615-623. 9. Kent KC, Donaldson MC, Attinger CE, et al. FemoropopUteal reconstruction for claudication. Arch Surg 1988;123,1196-1198. 10. CronenwenJL, Warner KG, lelenock GB, et aJ. Intermittent claudication. Arch 5urg 1984;119.430-436. 11. Coffman JD. Intermittent claudication and rest pain: physiologic concepts and therapeutic approaches. Prog Cardiovasc Dis 1979;163-72. 12. Raines J Larsen PE. Practical guidelines for establishing a clinical vascular laboratory. Cardiovascular Diseases, Bulletin of the Texas Heart Institute 1979; 1,9')-122. 13. Mannick JA. Current concepts in diagnostic methods: evaluation of chronic lower extremity ischemia. N Engl] Med 1983;309,841-844. 14. Raines JK, Carling RC, Buth J, et al. Vascular labora~ tory criteria for the management of peripheral vas~ cular disease of [he lower extremities. Surgery 1976; 79,21-29. 15. Ahn, 55, Rutherford RB, Becker GJ, et aL Reporting standards for lower extremity arterial endovascular procedures. J Vase 5urg 1993;17,110')-1107. 16. Owen RS, Carpenter JP, Baum RA, et al. Magnetic resonance imaging of angiographically occult runoff vessels in peripheral vascular occlusive disease. N Engl J Med 1992;326,1577-1581. 17. Yucel EK, Durmoulin CL, Waltman AC. MR angiog~ raphy of lower extremity alterial disease: preliminaJY experience. JM1U 1992;2,30')-309. 18. Edwards JM, Coldwell DM, Goldman Mi, el aL The role of duplex scanning in the selection of patients for lransluminal angioplasty. J Vas 5urg 1991;13,6974. 19. Monera GL, Strandness DE. Peripheral arterial du~ plex scanning. J Clin Ultrasound 1987;15,645-651.

P22

1,30 p.m. Nuts and Bolts of MRA James Carr, MD

Northwestern Memorial Hospital Chicago, Illinois

Atherosclerosis causes significant morbidity and mortal~ ity in the developed world and can affect any part of the circulation. Digital subtraction angiography (DSA) re~ mains the gold standard for evaluating the vasculature; however, this is an invasive investigation with welldescribed, potentially lethal side effects. As a result noninvasive methods for imaging blood vessels have been developed. Real~time Doppler ultrasound is useful for assessing superficial vessels (eg, carotid bifurcation, by~ pass grafts). However, it is operator dependent and has limited use in assessing deep vessels. cr angiography (CTA), panicularly using newer generation multidetector scanners, has recently emerged as a promising technique for evaluating the vasculature. However, CTA involves Significant radiatio~ exposure and injection of potentially nephrotoxic iodine-based contrast media. MR an~ giography (MRA) has developed as a useful noninvasive melhod for evaluating the circulation and has the advan~ rage of producing images similar to conventional angiog~ raphy. MR Angiography Techniques Two basic techniques have been utilized to image the vasculature with MR1: time-of-flight eTOp) imaging and contrast-enhanced MR angiography (ce~MRA). TOF Imaging TOF imaging relies on flowing blood to produce the MRI signal. Blood entering {he imaging volume has not been saturated by the RF pulse and therefore appears bright. Background tissue within the imaging volume is com~ pletely saturated and appears dark. The result is high contrast between flOWing blood and soft tissues. The advantage of this technique is that a contrast agent is not reqUired. Imaging times, however, are prohibitively long resulting in limited anatomic coverage and high susceptibility to motion artifacts. In addition, blood signaluniformity is variable in regions of turbulent flow and tortuous vessels.

Ce-MRA Ce-MRA has more recently taken over as the technique of choice for MR angiography. A spoiled gradient echo pulse sequence or FLASH (Fast Low Angle Shot) is used for MR angiography. This is a 3D fast Fourier transfOrm C3DFFT) gradient echo sequence with radiofrequency (RF) spoiling. FLASH will saturate spins whep the repetition time erR) becomes shorter than the T.1 of the tissues. As a result, Signal from background tissues is markedly reduced. When the T1 relaxation time is dramatically reduced using a paramagnetic substance such as Gadolinium~DTPA, much more Signal is available from the tissues. FollOWing an intravenous injection of

Gadolinium, only the arteries will have high signal when scanning occurs during the arterial phase of enhancement. The flip angle, Q, is chosen to optimize the blood signal. With FLASH-type sequences, maximal signal intensity occurs at a specinc flip angle, known as the Ernst angle QE' which is defined as: Q E = COS-I (TRlTI)

(1

With Gadolinium-enhanced MRA, optimal blood signal intensity rypically occurs at a flip angle of 25-30° The scan time, Ts ' for the MRA sequence is defined as: Ts = TR X Np X Ns Np and Ns denote the number of partitions and phase-encoding steps, respectively. Np and Ns define through-plane and in-plane spatial resolution. The more spatial resolution required, the longer the Ts ' The TR depends on the gradient hardware. With recent advances in gradient technology, much shorter repetition times (TRs) can be achieved. This has resulted in a marked reduction in imaging times allowing high-resolution ce-MRA to be carried out in a single breath-hold. TRs less than 2 msec can now be achieved using modern scanners. The short TR can be used to improve either spatial or temporal resolution. Minor sacrifices in through-plane and in-plane resolution can be made to improve temporal resolution. This is useful in areas with rapid arteriovenous transit allowing imaging to occur in the purely arterial phase of contrast enhancement. A sequence with an ultrashort TR results in a higher bandwidth of the readout Signal. This results in a lower signal to noise ratio (SNR). Increasing the bandwidth by a factor of 2 reduces the SNR by the square root of 2. Because acquisition times are reduced, Signal averaging can be used to increase the SNR. With newer software platforms, the bandwidth of the pulse sequence can be adjusted at the user interface. In general, the shortest TR possible is chosen for ce-MRA. With 3DFFT imaging techniques, a 3D k-space matrix is measured during the data acquisition. During every TR interval, a line of k space is acquired until the complete 3D k space matrix is filled. The center of k space defines image contrast while the periphery defines detail resolution of the image. In order to reduce the scanning time further, only a portion of k-space is actually acquired. Using different techniques, the remainder of k space is interpolated with only minor reductions in spatial resolution. The center of k space, which defines image contrast, is usually acquired first, typically in the first half of the scan time. Therefore, it is important that contrast be present within the arteries to be imaged when scanning starts. Because of this, the exact arrival of contrast in the arteries is determined prior to scanning. A number of different methods have been utilized to measure the contrast transit time. These include a timing bolus or flu oro-prep acquisition, which are carried out before the ce-MRA.

It is important that the 3D volume is positioned exactly over the region of interest, usually in a coronal orientation. A volume thickness of 80-1 OOmm is typically chosen, depending on the size of the patient, and must be of sufficient size so as to prevent wrap-around altifact. If the patient is very large, phase oversampling can be used to prevent wrap-around artifact. Approximately 80 partitions are used, depending on the required through-plane resolution. The field-of-view should also be optimized and rectangulated as much as possible as determined by the size of the patient. This will help keep the acquisition time to a mjnimum. As mentioned already, increasing the number of phase-encoding lines and partitions will improve spatial resolution but will also lengthen the imaging time. In addition, smaller voxel sizes will result in a reduced SNR. A precontrast 3D set is obtained first. The GadOlinium is injected typically at 2-3 rnVsec. The length of the contrast infuSion should approximate the acquisition time per 3D set. One or two post-contrast sets are then acquired. Breath-holding is utilized during all acquisitions to prevent respiratOlY motion artifact. Subtracted sets are obtained by subtracting the pre from the post. Different image postprocessing methods can be used to display the results including maximum intensity projection (MIP) and volume-rendering (VR) techruques. It is important also to evaluate the raw 3D data by directly examining the partitions or using multiplanar reformatting (MPR). The shorter TRs now achievable can be used to improve either spatial or temporal resolution. In general, conventional ce-MRA is implemented as a high spatial resolution technique in order to depict as much detail as possible. In some situations where demonstration of high-flow vascular lesions is important, spatial resolution can be sacrificed to improve temporal resolution. Ce-MRA has the advantage of producing images, which are similar to conventional angiography. Furthermore, the blood signal intensiry is consiStently uniform being relatively unaffected by the dephasing effect of flow disturbances. mtrafast MR Angiography Techniques Although conventional ce-MRA can prOVide detailed images of the vasculature with high spatial resolution, it does not illustrate how high-flow vascular abnormalities change with respect to time. This is palticularly important in conditions such as dissections, arteriovenous shunts and subclavian steal syndromes. As a resu.lt, a number of new techniques have been developed recently that attempt to improve the temporal resolution of ce-MRA. As gradient strength improves, it is possible to achieve much shorter repetition times. As a result, 3D ce-MRA can be carried out with subsecond temporal resolution. The basic pulse sequence is a 3D gradient echo acquisition, similar to conventional ce-MRA. Typical scanning parameters for subsecond 3D ce-MRA are as P23

follows : TR 1.6 msec; TE 0.8 msec; flip angle 25°; slab thickness 80 mm; 4 partitions (8 with sinc-interpolation); matrix size 256; voxel size 2.0 X 1.5 X 10 mm. Six millimeters of Gadolinium are injected at 6 mllsec via an 18-gauge cannula placed in an antecubital vein. The contrast injection and MR acquisition are started simultaneously and patients are asked to breath-hold in inspiration. Approximately 24 3D volumes are typically acquired in a single breath-hold. The first 3D set serves as a mask and subtraction occurs in-line. Maximum intensity projection (MIP) images are produced automatically. The entire series can be viewed as a cine loop with a frame time of 900 msec. In order to reduce the acquisition time per 3D volume to a minimum, through-plane resolution is sacrificed resulting in near-projectional MR angiography. This does not appear to interfere with image quality. The main objective is to depict high-flow vascular abnormalities in the plane of the scan with as high temporal resolution as possible and, as a result, through-plane information is less important. Above all, because of the light contrast load, subsecond ce-MRA can be repeated a number of times in combination with conventional ce-MRA in order to provide a comprehensive assessment of a vascular abnormality. The speed at whi;::h ce-MRA is carried out is primarily limited by the time it takes to acquire phase-encoding steps. This, in turn, is dependent on the performance of the gradient hardware. As gradient technology improves, this information is acquired more rapidly only at the expense of producing unwanted neuromuscular stimulation due to rapid gradient switching. A number of techniques have evolved that attempt to improve temporal resolution by changing the way data is acquired and processed. With the SMASH (simultaneous acquisition of spatial harmonics) and SENSE (sensitivity encoding) techniques, component coil signals in a radiofrequency coil array are used to partially encode spatial information by substituting for phase-encoding gradient steps that have been omitted. This allows some of the MR image to be acquired in parallel taking some of the strain off the gradients. Up to 4 fold accelerations have been demonstrated in humans . With 3D TRICKS (timeresolved imaging of contrast kinetics), the high spatial frequencies are sampled less frequently than the low spatial frequencies. As a result, high contrast information is preferentially acquired. Data sharing and temporal interpolation are also employed to further improve temporal resolution. All of these newly developed techniques can be combined with conventional 3D ce-MRA pulse sequences and have the potential to produce unprecedented improvements in temporal resolution.

Future Developments As gradient technology advances, scanning speed will increase resulting in much faster acquisition times. Software developments will produce more rapid data processing resulting in near real-time depiction of vessels. It may become possible to image the vasculature from P24

head to toe in a single study follOWing a single injection of contrast medium. In addition, newer intravascular contrast agents that remain within the circulation for longer periods of time will allow multiple acquisitions to take place without injecting additional contrast medium. The development of dedicated surface receiver coils will provide images with higher resolution and contrast. As scanner speed increases, it will become possible to carry out various interventional techniques on open scanners using real-time techniques.

Suggested Reading 1. Heiserman JE, Drayer BP , Fram EK, et aI. Carotid artery stenosis: clinical efficacy of two-dimensional time-of-flight MR angiography. Radiology 1992;182: 761-768. 2. Litt AW, Eidelman EM, Pinto RS, et al. Diagnosis of carotid artery stenosis: comparison of 2DIT time-offlight MR angiography with contrast angiography in 50 patients. AJRAMJ RoentgenoI1991;156:611-616. 3. Creasy J, Price R, Presbrey T, et al. Gadoliniumenhanced MR angiography. Radiology 1990;175: 280-283. 4. Prince M. Gadolinium-enhanced MR aortography. Radiology 1994;191:155-164. 5. Mitchell DG. MRI Principles. Philadelphia: WB Saunders, 1999. 6. Finn JP. Physics of MR Imaging. MR Imaging Clin North Am 1999;7:607-795. 7. Kim JK, Farb RI, GA W. Test bolus examination in the carotid artery at dynamic gadolinium-enhanced MR angiography. Radiology 1998;206:283-289. 8. Huston J, Fain 5B, Riederer 5J, et al. Carotid arteries:

maximizing arterial to venous contrast in fluoroscopically triggered contrast-enhanced MR angiography with elliptic centric view ordering. Radiology 1999; 211:265-273. 9. Levy RA, MR P. Arterial-phase three-dimensional contrast-enhanced MR angiography of the carotid arteries. AJR Am J Roentgenol 1996;167:211-215. 10. Carr J, McCalthy R, Laub G, et al. Subsecond, contrast-enhanced 3D MR angiography: a new technique for dynamic imaging of the vasculature. Proceedings of the 9th meeting of the International Society for Magnetic Resonance in Medicine 2001; 302. 11. Sodickson DWM. Simultaneous acquisition of spatial harmonics (SMASH): fast imaging with radiofrequency coil arrays. Magn Reson Med 1997;38:591603. 12. Sodickson D, McKenzie C, Li W, et al. Contrastenhanced 3D MR angiography with simultaneous acquisition of spatial harmonics: a pilot study. Radiology 2000;217:284-289.

13. Weiger M, Pruessmann K, Kassner K, et al. Contrastenhanced 3D MRA using SENSE. ] Magn Reson Imaging 2000;12:671-677. 14. Weiger M, Pruessmann K. Cardiac real-time imaging using SENSE. Magn Reson Med 2000;43:177-184. 15. Korosec FR, Grist TM, Mistretta CA. Time-resolved contrast-enhanced 3D MR angiography. Magn Reson Med 1996;36:345-351. 2:10 p.m. Imaging of Aortic Pathology MR lmaging Neil M. Rofsky, MD Beth Israel Deaconess Medical Center Boston, Massachusetts

Techniques Black Blood Strategies Cardiac triggered SE black-blood imaging is the "backbone" in the MR evaluation of the aorta. The R-wave acquisition at a constant spike is used to trigger image acąuisition point in the cardiac cycle "freezing" the heart and mediastinal structures. Saturation bands placed just above and below the extreme slices will minimize flow related enhancement (1). acquisitions have been possible by using echo Faster acąuisitions techniques (FSE or TSE). A newer modifitrain imaging techniąues cation that approach has been introduced providing exacquisitions (2). A halftremely rapid "black blood" acąuisitions Fourier single shot turbo spin-echo (HASTE) pulse sequence null seąuence that uses this double inversion pulse to nulJ Signal of blood is a particularly efficient techniąue technique to the signal study the aorta and can obviate the need for breath holding. A magnetization prepared gradient echo sequence can also achieve a black-blood image with T1ąuence weighted image contrast (3). Bright Blood Strategies Gradient echo flow sensitive imaging.-Gradient echo techniques can generate bright blood images and be

pared with breath-hold techniąues; techniques; the latter reąuire require high performance gradient systems. For breath hold imaging we use a spoiled 3D gradient echo sequence seąuence with [TRITE/flip angleJ; anglel; the following parameters: 3-5/2/50°, [TRlTE/flip 8-12 cm slab; 30-42 partitions; matrix size 128 X 256; acquisirectangular field-of-view (2:350 cm); and one acąuisi­ tion. The zero fill option results in smaller pixel sizes, which reduces partial volume artifacts and eliminates a off axis striped appearance that can be encountered on offaxis maximum intensity projections (7). Gadolinium can be administered intravenously (0.10.2 mmol/kg) mmollkg) as a bolus initiated 10 sec before data acquisition Ol' or as a slow continuous infusion (8). To acąuisition ensure that the peak arterial enhancement coincides acquisition of the central lines of K-space (rewith the acąuisition maximal contrast) we routinely use a "test dose" gion of maximaI Icc of gadolinium, with a power injector, to determine of 1cc the precise time to peak enhancement (9). Newer approaches include automated detection of the contrast bolus (10), MR "fluoroscopic" imaging of the acquisition bolus and subsequent subseąuent user initiation of the acąuisition (11) and very rapid temporal resolution strategies to ensure dut an appropriate window of contrast enhanceFurther improvements in ment is routinely captured (12). Furthel' gradient performance allow for the use of 3D sequences seąuences ;0;2; such rapid imaging can acąuire acquire data sets with a TR :=;2; every 4 seconds and will likely obviate the need for timing strategies. Phase contrast techniques can provide information about the direction and velocity of flow. These rely upon the position dependent phase shifts that occur when protons are exposed to magnetic gradients (13). Maximum intensity projections (MIPs) are most often used to display the Gd-enhanced aorta. However, it unique information is often should be remembered that uniąue only detected when evaluating the source (cross-sectional) data.

acqUired acąuired

at different phases of the cardiac cycle for dynamiC fashion. Flow is indicated by changes in signal dynamic intensity through the cardiac cycle on the cine display. useful for demonstrating flow, distinguishing flow This is usefuJ from thrombus, and showing the altered flow at the aortic valve as occurs with aortic valve stenosis or insufficiency. Gd-enhanced MR Angiography.-Gadolinium-enmulti planar reconstruction alhanced 3D studies offer multiplanar plane. In addition to lows for assessment in almost any pIane. required. A offering time efficiency, ECG triggering is not reąuired. acquisition with a large FOV can encompass the sagittal acąuisition entire aorta and is beneficial for studying the branch vessels of the thoracic and abdominal aorta (4-6). The injection should be performed through a rightsided venous catheter whenever possible to avoid brachiocephalic vein enhancement, wh!ch can degrade the presentation of the arch vessels. Non breath-hold techniques with slower sequences niąues seąuences are less preferable com-

Aortic Aneurysm The vast majority of thoracic aortic aneurysms are atherosclerotic. A dilatation of all components of the aortic wall results in a true aneurysm and is thought to occur from a loss of strength of the media. The arch and descending aorta are most commonly affected. An aneurysm of the ascending aorta will still be most commonly due to atherosclerosis. However, other etiologies should be considered - cystic medial necrosis, dissection and luetic aortitis. In asymptomatic patients cross-sectional imaging is used to monitor the size of aneurysms, as risk of rupture is also related to size. A rapid change in size or a diameter 2:6 cm are usually indications for surgical repair. The size criterion for surgical repair may vary on an institutional basis. Consistency in the method of measurement is essential for the longitudinal follow-up of aneurysmal disease. The use ofaxial of axial impatients with aneurysmaI

P25