Pe d i a t r i c I m a g i n g • C l i n i c a l Pe r s p e c t i ve Young et al. MR Angiography of Congenital Cardiovascular Disease

FOCUS ON:

Downloaded from www.ajronline.org by 37.44.207.15 on 01/24/17 from IP address 37.44.207.15. Copyright ARRS. For personal use only; all rights reserved

Pediatric Imaging Clinical Perspective

Tips and Tricks for MR Angiography of Pediatric and Adult Congenital Cardiovascular Diseases

Phillip M. Young1 Kiaran P. McGee Matthew S. Pieper Larry A. Binkovitz Jane M. Matsumoto Amy B. Kolbe Thomas A. Foley Paul R. Julsrud

OBJECTIVE. The use of contrast-enhanced MR angiography (MRA) as an alternative to CT angiography or conventional angiography to assess pediatric and adult patients with cardiovascular diseases has the potential to significantly reduce patients’ lifetime exposure to ionizing radiation. However, imaging this group of patients can be challenging because of a number of factors, including small size, difficulty timing the contrast bolus to the territory of interest, and the presence of metallic susceptibility artifact resulting from stents or clips. CONCLUSION. We present some suggestions to overcome many of these obstacles to MRA in these patients, highlighted with illustrations from clinical cases.

Young PM, McGee KP, Pieper MS, et al.

ontrast-enhanced MR angiography (MRA) is an attractive technique for noninvasive assessment of patients with pediatric and adult congenital cardiovascular diseases. The lack of ionizing radiation is a benefit of MRA in distinction to CT angiography (CTA) or conventional angiography for younger patients, because many such patients are serially imaged and will require lifetime follow-up, resulting in high cumulative radiation exposure. In addition, the ability to obtain a 3D dataset with wide anatomic coverage is an advantage of MRA over both conventional angiography and vascular ultrasound. Nevertheless, MRA can be challenging in these patients because of a number of specific issues. Younger patients often will not tolerate the MRA examination unless it is performed under general anesthesia, which entails its own risks [1]. In small children, obtaining adequate spatial resolution to depict cardiovascular pathologic abnormalities can be difficult. Smaller and younger patients are also difficult to image with MRA because it can be difficult to accurately time the contrast bolus to the territory of interest because of small bolus volume and rapid circulation time. Even in older children and adults who can be given larger contrast boluses, properly timing the scan delay from injection to highlight cardiovascular abnormalities can be difficult in the presence of shunting or postoperative anatomy, such as a Fontan circulation. Other challenges include

Keywords: congenital cardiovascular disease, CT angiography, MR angiography DOI:10.2214/AJR.12.9632 Received July 13, 2012; accepted after revision December 4, 2012. 1

All authors: Department of Radiology, Mayo Clinic, 200 First St SW, Rochester, MN 55905. Address correspondence to P. M. Young ([email protected]).

AJR 2013; 200:980–988 0361–803X/13/2005–980 © American Roentgen Ray Society

980

C

the presence of implanted metallic stents and surgical clips. Existing and evolving MRA techniques have the potential to help overcome many of these limitations. Judicious use of parallel imaging, blood-pool contrast agents, and other specialized techniques can improve image quality, spatial resolution, and overall diagnostic utility significantly over conventional approaches. With proper attention to technique, MRA is a powerful tool that can successfully image a wide range of cardiovascular abnormalities. We will present some “tips and tricks” we use in clinical practice to approach imaging problems, with a focus on state-of-the-art techniques available on newer scanners. Parallel Imaging Parallel imaging techniques intentionally undersample k-space, decreasing the amount of time required to acquire a given imaging volume. Briefly, these techniques all exploit coil geometry and the sensitivities of individual elements in the coil array for different regions of the imaged volume to extrapolate spatial information in the image. By doing this, phase encoding can be undersampled, and the spatial aliasing that would normally occur can be “unwrapped” using coil sensitivity information, which is obtained either before [2, 3] or during [4] the diagnostic scan. For 3D acquisitions such as the 3D spoiled gradient-echo sequences used for MRA, there are two phase-encoding directions; thus, 2D acceleration techniques can

AJR:200, May 2013

Downloaded from www.ajronline.org by 37.44.207.15 on 01/24/17 from IP address 37.44.207.15. Copyright ARRS. For personal use only; all rights reserved

MR Angiography of Congenital Cardiovascular Disease be applied given sufficient coil elements [5]. The resulting shortened duration of sampling required for a given scan prescription can be used in several ways: to shorten scan time for the same scan prescription [6], to improve spatial resolution in the same volume and scan time, or to increase the volume of scan coverage for a given scan time and spatial resolution [7]. Although parallel imaging techniques generally result in spatially varying signal-to-noise ratio (SNR) loss in an MRI, MRA techniques may experience a paradoxical improvement in SNR for a given scan prescription. This occurs because a faster scan can maximize SNR for a vascular territory because of the temporally varying gadolinium concentration in the vessel of interest [8]. The FOV should always be prescribed fully outside the patient’s skin in any direction for which parallel imaging is being used; otherwise, significant artifacts can result. In general, parallel imaging techniques can be tailored by the technologist to the specific clinical needs of the patient being imaged—faster scan time, higher resolution, or increased coverage—and result in excellentquality diagnostic images. Timing a Very Small Contrast Bolus Gadolinium-based contrast agents are usually dosed using a weight-based algorithm of 0.1–0.2 mmol/kg. In older children and adults, a small-volume (1 or 2 mL) timing bolus can be used to determine peak opacification of the vascular territory of interest, and that information can be used to determine the optimal delay after injection to begin the scan. However, in smaller children, the volume of a weight-based dose may be as small as 1 or 2 mL, and a timing bolus cannot be performed. “Fluoro triggering” is an option [9, 10], but in our hands it is too unreliable, especially between technologists with different levels of experience, and especially in this patient population. Time-resolved imaging with view-sharing techniques, such as keyhole [11], time-resolved imaging of contrast kinetics [12], time-resolved echo-shared angiographic technique or time-resolved angiography with interleaved stochastic trajectories [13], and Cartesian acquisition with projectionlike reconstruction [14], can be used with success [15]. However, these techniques are prone to significant artifacts because of view sharing and parallel imaging. These artifacts are accentuated in the presence of motion, such as that caused by respiration, cardiac pulsation, and bowel peristalsis, which can be

problematic in the chest, abdomen, and pelvis. View-shared techniques, as they are usually used, also can generate massive numbers of images, which are difficult and time consuming to review, particularly if both subtracted and unsubtracted images are constructed. An alternative approach is to obtain timeresolved non–view-shared MRA. With multichannel coils and 2D parallel imaging techniques, good-quality images can be obtained with excellent spatial resolution. Although the temporal resolution of this approach suffers in comparison with view sharing, the temporal resolution is adequate for many clinical applications and often has image quality superior to that of view-shared techniques (Fig. 1). In patients who are under general anesthesia, our approach is to have the anesthesiologist hyperventilate the paralyzed patient, then suspend respiration for 90 seconds. We prescribe a fiveor six-phase MRA with 2D autocalibrating reconstruction for Cartesian imaging acceleration, with each phase lasting 15–20 seconds and with automatic subtraction of a mask phase. The technologist or nurse quickly injects gadolinium-based contrast agent and flushes it with a three-way stopcock after the mask phase, and the resulting images are usually of very high diagnostic quality (Figs. 2 and 3). Contrast Agents Gadolinium-based contrast agents have some important differences in their imaging properties. The use of nearly all gadoliniumbased contrast agents for contrast-enhanced MRA is off-label according to the U.S. Food and Drug Administration (FDA) for almost all clinical indications. However, all FDA-approved agents have been successfully used for MRA, and the use of gadolinium-based agents for contrast-enhanced MRA is a widespread and widely accepted practice. It is important to note, however, that there are potentially significant differences between the characteristics of some of the various approved agents. Two agents in particular have potential benefits for MRA in our experience. Gadobenate dimeglumine (Multihance, Bracco) generally has a higher intravascular signal than do conventional agents because of its capacity for weak transient interaction with serum albumin, which further quickens the rate of T1 relaxation when compared with standard gadolinium chelates [16–18]. The result is brighter intravascular contrast and longer persistence of contrast in the vascular tree. The effects of greater T1 relaxation and longer intravascular residence time from albumin

interaction are even more pronounced with gadofosveset trisodium (Ablavar, Lantheus Billerica, MA; known in Europe and Asia as Vasovist). This agent is approved for aortoiliac MRA by the FDA. The albumin-binding properties of this agent make steady-state “blood-pool phase” imaging possible, in some cases for up to and beyond 1 hour after contrast injection. By using parallel imaging techniques to improve anatomic coverage and resolution, the blood-pool phase can be exploited to obtain very detailed anatomic images of cardiovascular anatomy and can be particularly useful to image territories in which optimal contrast timing is difficult, such as cavopulmonary anastomoses. We have found that a combination of dynamic and blood-pool phase breath-hold techniques with 2D accelerated MRA can be particularly helpful for imaging the entire chest in patients with Fontan circulation, aortic coarctation, and other congenital cardiovascular diseases (Figs. 4–6). When injecting contrast agent into any patient with congenital heart disease, it is prudent to remind the responsible personnel to be extremely cautious to avoid injecting air bubbles into the venous circulation, because of the potential risk of right-to-left shunting and paradoxical air emboli. Venous Imaging and Vascular Malformations The blood-pool phase of gadofosveset administration is also an excellent technique for venous imaging. This is particularly advantageous in settings where imaging with and without provocative maneuvers is required, such as popliteal entrapment syndrome [19] or thoracic outlet syndrome [20]. The use of blood-pool contrast agents eliminates the need for reinjection and coordination of bolus timing with exertion or positioning (Figs. 7 and 8). Because the scan times for extremity imaging are not limited by breath-holding, images with extremely high spatial resolution can be obtained with longer scan times (Fig. 9). When imaging vascular malformations in the periphery, we usually use view-shared time-resolved MRA with Cartesian acquisition with projectionlike reconstruction or time-resolved imaging of contrast kinetics (Fig. 10) to maximize temporal resolution. However, if imaging in the abdomen or pelvis, we sometimes use non–view-shared time-resolved MRA to eliminate respiratory motion artifacts when a lesion is known or suspected to be low flow (Fig. 11).

AJR:200, May 2013 981

Downloaded from www.ajronline.org by 37.44.207.15 on 01/24/17 from IP address 37.44.207.15. Copyright ARRS. For personal use only; all rights reserved

Young et al. Imaging in the Presence of Metallic Stents When patients have had metallic stents placed for arterial or venous stenosis, the resulting susceptibility artifact frequently limits assessment of the stent lumen on MRI. Although the artifact cannot be eliminated, some modifications to scan technique can minimize the artifact and often lead to goodquality images. The receiver bandwidth can be increased to the maximum value (250 kHz in most cases) to limit the influence of susceptibility on adjacent tissue, and the flip angle can be increased to the maximum allowable value to overcome the Faraday cage effect. Although these measures will help minimize susceptibility, the improvements come at a cost of decreased SNR. It may therefore be beneficial to image in the bloodpool phase with both standard technique (for most of the vascular tree) and altered bandwidth and flip angle (for the stented component) to provide the highest overall diagnostic quality (Fig. 12). Sequential Versus Elliptic Centric Readout Most MRA sequences are built by default to use elliptic centric readout, in which the center of k-space (low spatial frequencies) is filled at the beginning of the scan, and the remainder of k-space is filled from the center outward. Because most of the information in the examination exists in the center part of k-space, this approach is combined with rapid injection rates (≥ 3 mL/s) to allow timing of a compact contrast bolus to the vessel of interest at the precise time the center of k-space is acquired. However, this approach may not work well when longer scan times are used, because the rapid washout of the compact contrast bolus leaves little signal in the vessel when the periphery of k-space is acquired at the end of the scan. For example, injecting a 15-mL contrast bolus at 3 mL/s means the injection lasts 5 seconds. If the patient can suspend respiration for a 30-second scan, there is a mismatch between the duration of the contrast bolus and the duration of the scan, with the contrast agent washed out when the periphery of k-space is acquired. The result is a loss of the fine edge detail represented by the periphery of k-space—in effect, a blurry examination. We tend to favor very-high-resolution matrices, thin slices, and extended anatomic coverage, which tends to lengthen the scan time beyond the duration of a contrast bo-

982

lus injection, as described already. For example, even in smaller patients (FOV < 30 cm) we will frequently use matrices of 320– 384 × 160–256 (frequency times phase) and slice thicknesses of 1.8–3 mm, resulting in scan times of 20–30 seconds. In larger patients (FOV ≥ 40 cm), we will use matrices of 384–512 × 256–512 and slice thickness of 0.6–3 mm. To maximize intravascular signal during a relatively longer scan time, we frequently use sequential readout (in which kspace is filled from top to bottom, and the center is filled halfway through the scan), double-dose gadolinium administration (0.2 mmol/kg), and “stretch” the contrast bolus by injecting at a slow rate (1 or 1.5 mL/s). We use a 1- or 2-mL timing bolus to calculate the timing of peak intravascular concentration. Because the peak gadolinium concentration must be timed to the center of k-space acquisition, half of the scan time must be subtracted from the peak time, and then the technologist adds the full contrast bolus volume divided by twice the flow rate (to correct for the difference in volume between the test bolus and the actual contrast bolus). Thus, the timing equation is as follows: scan delay = peak time from bolus timing − (scan time / 2) + [contrast volume / (2 × flow rate)]. Although the process of timing the bolus is more difficult, the resulting images are often higher in spatial resolution and better in image quality than can be achieved with elliptic centric contrast techniques. These can be used to improve vascular assessment when injecting contrast agent for other purposes, such as for tumor staging and preoperative planning (Fig. 13). All of the images shown in this article were obtained with sequential readout except for Figures 1A and 10. Visual Communication of Findings With Postprocessed Images One of the major barriers to clinical use of MRA for cardiovascular disease is that the images and techniques are even more difficult for clinicians to understand than for radiologists. MRA examinations for cardiovascular diseases often contain numerous pulse sequences, which are confusingly named, and sometimes thousands of images that cannot be rapidly reviewed. In addition to a concise report, a small set of images containing relevant findings can help communicate relevant information to referring physicians and patients. Postprocessing techniques, such as maximum intensity projections and volumerendered images, can beautifully display clini-

cally relevant findings in a more intuitive format for the nonradiologist. Conclusion MRA is a powerful tool for investigating pediatric and adult congenital cardiovascular diseases. Although there are some inherent challenges to imaging this patient population, attention to detail and the use of existing and developing MRA techniques can overcome many of the limitations to cardiovascular assessment with MRA. References 1. Serafini G, Zadra N. Anaesthesia for MRI in the paediatric patient. Curr Opin Anaesthesiol 2008; 21:499–503 2. Sodickson DK, Manning WJ. Simultaneous acquisition of spatial harmonics (SMASH): fast imaging with radiofrequency coil arrays. Magn Reson Med 1997; 38:591–603 3. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE: sensitivity encoding for fast MRI. Magn Reson Med 1999; 42:952–962 4. Griswold MA, Jakob PM, Heidemann RM, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 2002; 47:1202–1210 5. Fenchel M, Nael K, Ruehm S, Finn JP, Miller S, Laub G. Isotropic high spatial resolution magnetic resonance angiography of the supra-aortic arteries using two-dimensional parallel imaging (iPAT2) at 3 Tesla; a feasibility study. Invest Radiol 2006; 41:545–552 6. Sutter R, Heilmaier C, Lutz AM, Weishaupt D, Seifert B, Willmann JK. MR angiography with parallel acquisition for assessment of the visceral arteries: comparison with conventional MR angiography and 63-detector-row computed tomography. Eur Radiol 2009; 19:2679–2688 7. Lum DP, Busse RF, Francois CJ, et al. Increased volume of coverage for abdominal contrast-enhanced MR angiography with two-dimensional autocalibrating parallel imaging: initial experience at 3.0 Tesla. J Magn Reson Imaging 2009; 30:1093–1100 8. Riederer SJ, Hu HH, Kruger DG, Haider CR, Campeau NG, Huston J. Intrinsic signal amplification in the application of 2D SENSE parallel imaging to 3D contrast-enhanced elliptical centric MRA and MRV. Magn Reson Med 2007; 58:855–864 9. Luccichenti G, Cademartiri F, Ugolotti U, Marchesi G, Pavone P. Magnetic resonance angiography with elliptical ordering and peripheral MR angiography fluoroscopic triggering of the renal arteries. Radiol Med (Torino) 2003; 105:42–47 10. Riederer SJ, Bernstein MA, Breen JF, et al. Threedimensional contrast-enhanced MR angiography with realtime fluoroscopic triggering: design specifications and technical reliability in 330 patient

AJR:200, May 2013

Downloaded from www.ajronline.org by 37.44.207.15 on 01/24/17 from IP address 37.44.207.15. Copyright ARRS. For personal use only; all rights reserved

MR Angiography of Congenital Cardiovascular Disease studies. Radiology 2000; 215:584–593 11. van Vaals JJ, Brummer ME, Dixon WT, et al. “Keyhole” method for accelerating imaging of contrast agent uptake. J Magn Reson Imaging 1993; 3:671–675 12. Korosec FR, Frayne R, Grist TM, Mistretta CA. Time-resolved contrast-enhanced 3D MR angiography. Magn Reson Med 1996; 36:345–351 13. Wang Y, Lee HM, Khilnani NM, et al. Boluschase MR digital subtraction angiography in the lower extremity. Radiology 1998; 207:263–269 14. Haider CR, Glockner JF, Stanson AW, Riederer SJ. Peripheral vasculature: high-temporal- and high-spatial-resolution three-dimensional contrastenhanced MR angiography. Radiology 2009;

253:831–843 15. Goo HW, Yang DH, Park IS, et al. Time-resolved three-dimensional contrast-enhanced magnetic resonance angiography in patients who have undergone a Fontan operation or bidirectional cavopulmonary connection: initial experience. J Magn Reson Imaging 2007; 25:727–736 16. Giesel FL, von Tengg-Kobligk H, Wilkinson ID, et al. Influence of human serum albumin on longitudinal and transverse relaxation rates (R1 and R2) of magnetic resonance contrast agents. Invest Radiol 2006; 41:222–228 17. Rohrer M, Bauer H, Mintorovitch J, et al. Comparison of magnetic properties of MRI contrast media solutions at different magnetic field strengths. In-

A

vest Radiol 2005; 40:715–724 18. Pintaske J, Martirosian P, Graf H, et al. Relaxivity of gadopentetate dimeglumine (Magnevist), gadobutrol (Gadovist), and gadobenate dimeglumine (MultiHance) in human blood plasma at 0.2, 1.5, and 3 Tesla. Invest Radiol 2006; 41:213–221 19. Beitzke D, Wolf F, Juelg G, Lammer J, Loewe C. Diagnosis of popliteal venous entrapment syndrome by magnetic resonance imaging using bloodpool contrast agents. Cardiovasc Intervent Radiol 2011; 34(suppl 2):S12–S16 20. Lewis M, Yanny S, Malcolm PN. Advantages of blood pool contrast agents in MR angiography: a pictorial review. J Med Imaging Radiat Oncol 2012; 56:187–191

B

Fig. 1—Girl with truncus arteriosus type 1 who underwent neonatal repair and multiple revisions of right ventricular-pulmonary artery conduit. A and B, Single coronal contrast-enhanced MR angiography (MRA) with time-resolved view-shared parallel imaging was obtained when patient was 11 years old (A), and single coronal non–view-shared 2D autocalibrating reconstruction for Cartesian imaging–accelerated MRA was obtained 3 years later when patient was 14 years old (B). Both images are from best phase of opacification at approximately same location. There is increased edge definition and depiction of vascular and nonvascular structures in non–view-shared image (arrows, B) and subjectively improved image quality when compared with earlier view-shared image (arrows, A). FOVs, matrices, and slice thicknesses were similar between two examinations.

A

B

Fig. 2—5-month-old girl with pulmonary sequestration. MR angiography (MRA) was performed to document arterial supply and venous drainage of sequestration before surgery. A and B, Thin-slab maximum-intensity-projection images from autosubtracted 2D autocalibrating reconstruction for Cartesian imaging–accelerated MRA were acquired during continuous 90-second breath-holding after hyperventilation and show systemic arterial supply (arrow, A) and pulmonary venous drainage (arrow, B) of sequestration on consecutive phases.

AJR:200, May 2013 983

Downloaded from www.ajronline.org by 37.44.207.15 on 01/24/17 from IP address 37.44.207.15. Copyright ARRS. For personal use only; all rights reserved

Young et al.

A

B

C Fig. 3—5-year-old girl with cutaneous vascular lesion. A, Photograph of lesion. B, MR angiography (MRA) was performed to show whether lesion was purely cutaneous, or whether it involved deeper structures. Volume-rendered image from single phase of dynamic subtracted 2D autocalibrating reconstruction for Cartesian imaging–accelerated MRA shows high fidelity to photograph (A). C, Axial oblique thin-slab volume rendering of subtracted images shows lesion confined to subcutaneous tissue. On biopsy, this was shown to be malignant kaposiform hemangioendothelioma, rare vascular skin cancer.

984

AJR:200, May 2013

Downloaded from www.ajronline.org by 37.44.207.15 on 01/24/17 from IP address 37.44.207.15. Copyright ARRS. For personal use only; all rights reserved

MR Angiography of Congenital Cardiovascular Disease

A

B

C

D

Fig. 4—8-year-old girl with history of native hypoplastic left heart syndrome and anomalous total pulmonary venous connection to coronary sinus. Patient underwent initial Norwood procedure and repair of pulmonary venous connection to coronary sinus at age 6 weeks. She underwent hemiFontan procedure at age 8 months, but eventually thrombosed left pulmonary arterial system. Because of cyanosis, modified Blalock-Taussig shunt was performed from innominate artery to right pulmonary artery. MR angiography (MRA) was performed to assess status of her venopulmonary connections and Blalock-Taussig shunt because of increasing cyanosis. Time-resolved non–view-shared 2D autocalibrating reconstruction for Cartesian imaging–accelerated MRA was performed with injection of gadofosveset. A and B, Maximum-intensity-projection images at sequential time points show gadolinium from right arm contrast agent injection traveling caudally through chest wall collaterals (arrow, A) rather than into right subclavian vein, superior vena cava, and pulmonary circulation. Systemic atrium and systemic circulation are opacified from contrast agent arriving from trans–chest wall and diaphragmatic collaterals (A) rather through pulmonary arterial and venous circulations (visualization of right-left shunt). At later time frame (B), venous collaterals in left chest opacify (arrows, B). C, Mask-subtracted thin-slab volume-rendered image in blood-pool phase shows occlusion of left innominate vein and stenosis of right innominate vein (arrows) as cause of superficial venous collaterals and increased right-to-left shunting. D, Coronal oblique blood-pool phase reformatted image shows moderate stenosis at distal anastomosis of right Blalock-Taussig shunt (arrow) as secondary cause.

Fig. 5—49-year-old woman with partial anomalous venous return (“scimitar syndrome”). Thin-slab maximum-intensity-projection image of bloodpool phase contrast-enhanced MR angiography shows draining vein entering inferior vena cava in supradiaphragmatic position jointly with right hepatic vein (arrow). Use of blood-pool contrast agent eliminates need for precise timing of contrast bolus in difficult location where pulmonary venous, hepatic venous, and vena caval opacification are all different.

AJR:200, May 2013 985

Downloaded from www.ajronline.org by 37.44.207.15 on 01/24/17 from IP address 37.44.207.15. Copyright ARRS. For personal use only; all rights reserved

Young et al. Fig. 6—25-year-old woman with newly discovered coarctation of aorta. Two-dimensional autocalibrating reconstruction for Cartesian imaging–accelerated MRI was used. Because patient was good breath holder, 30-second MR angiography was prescribed. Rather than shortening scan time, parallel imaging was used to extend anatomic coverage across entire chest while maintaining 2-mm slice thickness. A, Sagittal oblique thin-slab volume-rendered image shows severe aortic coarctation (arrow) with enlarged intercostal collateral vessels. B, Coronal subtracted volume-rendered image shows extensive intercostal artery collateralization across chest, which would not be visualized as well without extended coverage provided by parallel imaging.

A

B Fig. 7—16-year-old girl with recurrent effort-related left upper extremity thrombosis. A and B, After injection of blood-pool contrast agent through contralateral arm, thin-slab maximumintensity-projection images obtained with arms abducted (A) and adducted (B) show narrowing of left subclavian vein with shoulder abduction (arrow, A), which disappears with adduction (B), compatible with venous thoracic outlet syndrome.

A

B

A

B

Fig. 8—18-year-old woman with functional popliteal entrapment syndrome. A and B, Thick-slab maximum-intensity-projection images from blood-pool phase contrast-enhanced MR angiography with 2D autocalibrating reconstruction for Cartesian imaging acceleration show bilateral popliteal artery and vein occlusion (arrows, A) with resisted plantar flexion, which is not present at rest (B).

986

Fig. 9—29-year-old woman with Klippel-Trenaunay syndrome. Coronal 2D autocalibrating reconstruction for Cartesian imaging–accelerated MR angiography image of calves in blood-pool phase of contrast opacification shows excellent depiction of venous structures, with thrombus in dilated peroneal vein (solid arrow) and dilated superficial venous varicosities (dashed arrow) in right leg.

AJR:200, May 2013

Downloaded from www.ajronline.org by 37.44.207.15 on 01/24/17 from IP address 37.44.207.15. Copyright ARRS. For personal use only; all rights reserved

MR Angiography of Congenital Cardiovascular Disease Fig. 10—16-year-old boy with slow-flow vascular malformation in right thigh. Cartesian acquisition with projectionlike reconstruction time-resolved view-shared MR angiography (MRA) with 2D array spatial sensitivity-encoding technique acceleration was performed. A, Maximum-intensity-projection image of one time point shows small perforator arteries (arrows) supplying malformation. B, Delayed blood-pool phase high-resolution Cartesian acquisition with projectionlike reconstruction MRA shows lesion contained within right vastus medius muscle (arrow).

A

B

A

B

C

Fig. 11—4-year-old boy with slow-flow vascular malformation in left gluteus medius. A–C, To avoid deleterious effect of respiratory motion and bowel peristalsis on image quality seen with view-shared MR angiography (MRA), time-resolved breath-held non–view-shared MRA was obtained with 2D autocalibrating reconstruction for Cartesian imaging acceleration. Autosubtracted images at sequential time points show temporal and spatial pattern of contrast opacification in slow-flow vascular malformation (arrows, A–C).

A

B

Fig. 12—17-year-old girl with native hypoplastic right heart syndrome and previous bidirectional Glenn and extracardiac Fontan surgeries. Metallic stent had been placed in suprahepatic Fontan because of stenosis. MR angiography (MRA) was performed to assess systemic to pulmonary venous connections because of fatigue. A, Blood-pool phase 2D autocalibrating reconstruction for Cartesian imaging–accelerated MRA shows widely patent Glenn (dashed arrow) and Fontan (dotted arrow) anastomoses with right pulmonary artery, but lumen of stent (solid arrow) is not well seen because of metallic susceptibility. B, By maximizing receiver bandwidth and flip angle (increased in this case from 25° to 60°), lumen of stent is better imaged (arrow), although at expense of decreased signal-to-noise ratio.

AJR:200, May 2013 987

Downloaded from www.ajronline.org by 37.44.207.15 on 01/24/17 from IP address 37.44.207.15. Copyright ARRS. For personal use only; all rights reserved

Young et al.

A

B

C

Fig. 13—16-year-old girl with sarcoma of right leg. A–C, Subtracted maximum-intensity-projection MR angiography (MRA) (A) and thin-section nonsubtracted MRA (B) can be correlated with anatomic tumor imaging, such as this contrast-enhanced fat-suppressed spoiled gradient-echo image (C) to facilitate operative intervention.

988

AJR:200, May 2013