HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lapeyre, M. Articles by Luciani, A. Search for Related Content PubMed PubMed Citation Articles by Lapeyre, M. Articles by Luciani, A. Social Bookmarking                      What's this?

Assessment of Critical Limb Ischemia in Patients with Diabetes: Comparison of MR Angiography and Digital Subtraction Angiography

¹ Service de Radiologie et d'Imagerie Médicale, Centre Hospitalier Universitaire Henri Mondor, 51 Avenue du Mal. De Lattre de Tassigny, 94010 Creteil Cedex, France.
² Service de Chirurgie Vasculaire, Centre Hospitalier Universitaire Henri Mondor Hospital, 94010 Creteil Cedex, France.

Received July 14, 2004; accepted after revision December 17, 2004.

 
Address correspondence to H. Kobeiter.

Supported in part by a grant provided by the O.P.A.L. Foundation, Sanofi-Synthelabo, France.

Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
OBJECTIVE. The purpose of our study was to evaluate the diagnostic accuracy of hybrid MR angiography by comparison with digital subtraction angiography (DSA) in diabetic patients with critical limb ischemia.

SUBJECTS AND METHODS. Thirty-one patients prospectively underwent both hybrid MR angiography and DSA. The hybrid MR angiography study consisted of high-resolution MR angiography of a single calf and foot using a contrast-enhanced 3D gradient-echo volumetric interpolated breath-hold examination with surface coils, followed by three-station bolus chase MR angiography with a dedicated peripheral vascular coil. Two blinded reviewers separately analyzed maximum-intensity-projection hybrid MR angiograms and DSA images. The peripheral vessels were divided into 10 anatomic segments for review. The status of each segment was graded as normal, stenosis less than 50% in diameter, stenosis greater than 50%, or occluded. The sensitivity and specificity of hybrid MR angiography were determined using DSA as the gold standard. Treatment options were considered separately from the results of each examination.

RESULTS. Among 310 analyzed segments, the sensitivities of hybrid MR angiography for stenosis and occlusion were, respectively, 95% and 95% for reviewer 1 and 96% and 90% for reviewer 2. The specificities of hybrid MR angiography for stenosis and occlusion were, respectively, 98% and 98% for reviewer 1 and 98% and 99% for reviewer 2. In 25 patients (81%), the quality of bolus chase MR angiography images was insufficient to assess runoff arteries. All treatments proposed on the basis of DSA findings were endorsed by hybrid MR angiography findings. Eleven more treatments were formulated on the basis of hybrid MR angiography findings. Of these, four were due to overestimation of stenosis on MR angiography and seven were due to the detection of patent infrageniculate arteries on hybrid MR angiography that were not detected on DSA.

CONCLUSION. Hybrid MR angiography depicts runoff arteries not seen on DSA. Hybrid MR angiography may be useful for treatment planning in selected diabetic patients with critical limb ischemia.

Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Diabetic patients are four times more likely than the general population to develop peripheral artery disease [1]. Peripheral artery disease is also associated with numerous distal stenoses and occluded runoff vessels [2-5], with five times more diabetic patients than nondiabetic individuals developing critical limb ischemia. In diabetic patients, critical limb ischemia requires aggressive treatment such as distal bypass or complex endovascular interventions [6, 7]. Moreover, despite surgical revascularization, lower limb amputation is approximately five times more frequent in diabetic patients than in nondiabetic individuals [8]. Several studies have shown that the evaluation of peripheral artery disease is the main prognostic determinant for amputation [9, 10]. The location, length, and severity of stenoses and the patency of runoff vessels must be precisely assessed before planning revascularization procedures [10-12].

Intraarterial digital subtraction angiography (DSA) is still used as the benchmark for peripheral artery disease evaluation, although cross-sectional imaging (i.e., MR angiography, CT angiography, or sonography) has become more frequently used [13]. On DSA assessing the global vasculature, detailed delineation of the calf vessels may not be obtained when numerous stenoses are present. Several studies have shown that DSA may thus fail to show runoff vessels in patients with peripheral artery disease associated with diabetes [11, 12]. MR angiography has given promising results in peripheral artery disease, including in diabetic patients [12, 14-16]. High spatial resolution is needed because of the distal location of the arterial stenoses, and high temporal resolution is needed because venous signal contamination is more troublesome in patients with critical limb ischemia [17-19].

The aims of this study were to assess the diagnostic accuracy of a new MR angiography protocol designed to analyze both diabetes-specific lesions of infrageniculate vessels and the entire peripheral vascular tree in diabetic patients with single-limb critical ischemia, and to report the therapeutic implications of MR angiography findings.

Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Patients
Between February 2002 and March 2003, 91 long-term diabetic patients with suspected critical limb ischemia were referred to the department of vascular surgery at our institution. The inclusion criteria for this study were nonhealing ulceration or focal gangrene consistent with peripheral artery disease on physical examination by a vascular surgeon [20]. Twenty-one patients were not enrolled because, in the view of the vascular surgeon, trophic changes were related to neuropathy or infection. Additional exclusion criteria were prior below-knee amputation on the same side (n = 6), contraindication to MR angiography (pacemaker, n = 5; claustrophobia, n = 2; ferromagnetic material, n = 3), previous arterial stenting that could render MR angiography inconclusive (n = 5), the nonavailability of MRI within 10 days after the initial clinical examination (n = 16), allergy to iodinated contrast agents (n = 1), and refusal of DSA (n = 1). A total of 31 patients were included in this prospective study. MR angiography was performed first so that endovascular treatment could be performed during DSA.

The 31 patients (22 men and 9 women) ranged in age from 35 to 83 years (mean, 65 years; median, 65 years). Trophic changes were nonhealing ulceration in 16 patients and focal gangrene in 15 patients. All patients had diabetes mellitus, of type 2 in 23 cases (74%). Nineteen patients had serum creatinine levels greater than 120 µmol/L (normal upper range), and three patients were undergoing dialysis. Thirty-one limbs were evaluated (no bilateral trophic changes were included).

Informed consent was obtained from all the patients, and the study protocol was approved by our institutional review board.

MR Angiography
MRI was performed on a 1.5-T scanner (Symphony Quantum system, Siemens Medical Solutions) equipped with high-performance 30 mT/m gradients and authorizing a maximal slew rate of 125 T/m per second. Our peripheral hybrid MR angiography protocol consisted of two distinct acquisitions, assessing primarily the infrageniculate arteries and, secondarily, the entire peripheral arterial tree.

Infragenicular MR angiography—Patients were placed in the supine position, feet first. The symptomatic limb was positioned in the center of the magnet, lying on a spine array coil. To obtain maximal contact between the surface coils and the calf and ankle and to optimize the study of pedal vessels, the lower limb was positioned in maximal external rotation. The surface phased-array body coil was placed anteriorly on the symptomatic calf and ankle. The contralateral calf was placed outside the coil to avoid aliasing artifacts. The infragenicular arteries of the symptomatic limb were studied first with a 3D gradient-echo T1-weighted fast low-angle shot spoiled sequence called VIBE (volumetric interpolated breath-hold examination). In this sequence, the transverse magnetization signal is spoiled in the slice and readout directions. The signal is first acquired in the partition-encoding direction and the data are interpolated in the phase and slice directions. Interpolation directly increases temporal resolution: the sequence is faster than the usual 3D T1-weighted gradient-echo sequences. Combination of the 3D data set encoding in the slice selection-encoding direction with interpolation increases vessel-to-tissue contrast. The parameters used are summarized in Table 1. The matrix size along the feet-head readout direction was always 512. Fat saturation was used after localized manual shimming. Imaging was performed in the sagittal plane of the limb. A series obtained before contrast enhancement was used as a mask and for subtraction from subsequent images. Contrast medium (0.01 mmol/kg of gadoterate dimeglumine [Dotarem, Guerbet]) was administered at 3 mL/sec, followed by a 20-mL saline flush administered at the same rate. To optimize bolus tracking, we used CARE bolus software (Siemens). Axial 2D single-slice multiphase gradient-echo images were acquired at the level of the popliteal artery to determine the arrival of the contrast bolus. The acquisition rate was approximately one image per second. The VIBE sequence was launched when the bolus arrived in the popliteal artery (or in the collaterals in cases of popliteal artery thrombosis). The unenhanced 3D mask sequence was automatically subtracted from the 3D contrast-enhanced VIBE sequence. The examinations were interrupted after the VIBE studies, and the patient was taken to the waiting room for 60-90 min. VIBE postprocessing was standardized: angiograms of the subtracted images were created using the maximum-intensity-projection (MIP) algorithm, and a series of six MIP images was generated for each 15° rotational increment from right lateral to left lateral. Technicians generated all the images.

Aortic and lower limb MR angiography—Sixty to ninety minutes after infragenicular examination, three-station bolus chase MR angiography of both legs was performed with multiphase 3D fast spoiled gradient-recalled echo sequences using a dedicated peripheral vascular coil, body and phased-array coils, and a moving table. Three stations (pelvis, thigh, and calf) were studied. Only the first station was studied with the patient in breath-hold. The parameters (field of view, matrix size, partition number, and slice thickness) at each station were optimized to permit rapid scanning while maximizing resolution and are summarized in Table 1. An un-enhanced acquisition was obtained for each station to serve as a mask for further examination. Real-time bolus tracking was performed using the CARE bolus system. An axial 2D single-slice image was placed at the level of the juxtarenal aorta to determine the arrival of the contrast bolus. The slice was repeated each second after injection of 0.02 mmol/kg of gadoterate dimeglumine (Dotarem). The injection parameters were as follows: phase 1, 0.015 mmol/kg at 2 mL/sec; phase 2, 0.005 mmol/kg at 0.8 mL sec; and then a flush with 30 mL of saline at 0.8 mL sec. Three-station stepping-table MR angiography was performed when enhancement of the aorta was obtained. The unenhanced 3D mask sequence was automatically subtracted from the 3D contrast-enhanced sequence for each station. Posttreatment images were processed by technicians. Angiograms of the subtracted images were created using the MIP algorithm. For each station, a series of three MIP images of each leg was generated using three projections, namely, left anterior oblique (30°), posteroanterior, and right anterior oblique (30°). All MR angiography studies were printed on hardcopy films. Window level settings for all MIP images were adjusted to maximize arterial contrast and minimize the background signal. Postprocessing was done by technicians.

Digital Subtraction Angiography
All DSA examinations were performed by experienced angiographers but not by the reviewers. The angiographers used a digital angiographic unit (Diagnostar, Philips Medical Systems; or Series 9800, OEC Medical Systems). DSA was always performed within 72 hr after MR angiography. In 17 patients, an anterograde common femoral artery approach was used in single-leg DSA examinations. A 4-French introducer sheath was positioned in the common femoral artery. Fourteen patients underwent selective DSA with crossover anterograde catheterization of the external iliac artery. Whatever the approach, multiple DSA images were obtained by injecting iodixanol 320 (Visipaque, Guerbet) with a power injector (Mark V, Medrad).

The DSA examinations consisted of anteroposterior overlapping evaluations of the lower extremities, except for the pedal and ankle stations, where a 15-30° external rotation of the foot was used. The required number of stations was adjusted to the field of view of the angiographic unit (Diagnostar, 40 cm; OEC 9800, 31 cm) to image the entire arterial vascular tree from the common femoral artery to the pedal arteries. The ankle and pedal stations were studied with a 30-cm field of view on the Diagnostar. The patient was repositioned for each imaging station, starting from the common femoral artery to the pedal arteries.

The injection parameters were as follows: volume, 8 mL and flow rate, 4 mL/sec for the above-knee-level sequences; and volume, 12 mL and flow rate, 5 mL/sec for the below-knee-level sequences. For each station, the frame rate was four images per second, and images were acquired until complete lack of visibility of the most distal artery of the station. All criteria of optimized angiography described by Gates and Hartnell [21] were met, except the use of drug vasodilation. Endovascular therapeutic procedures were performed at the same time in 22 of the 31 patients.

Images were stored on a hard disk before filming. For each acquisition, postprocessing was performed by the operator who performed DSA. For each station, the first image of a series was selected as a mask. The unenhanced mask image was automatically subtracted from the contrast-enhanced images for each station. For each station, the final angiogram was obtained by adding a selection of two to three subtracted images. Subtracted images were selected by the DSA operator to obtain homogeneous opacification of the entire arterial tree along the feet-head axis. Images were printed on hard copy (four images from one film).

Image Analysis
For each MR angiography and DSA study, the patient's name was obscured by tape, and a study number was assigned to each examination. The MR angiograms and DSA images were randomly organized and were placed so that interpretation was performed in different orders and on separate days. Two reviewers, an interventional radiologist with 13 years' experience (reviewer 1) and a vascular surgeon with 14 years' experience (reviewer 2), independently reviewed the MR angiograms and DSA images. They reviewed the films and formulated treatment plans separately on the basis of the MR angiography and DSA images.

For the comparison of infragenicular MR angiography analysis, third-step bolus chase MR angiography findings and VIBE sequence calf study findings were randomly organized and the images were placed so that interpretations were performed in different orders.

The same form was completed by the reviewers for both MR angiography and DSA. The quality of all examinations was graded on a 3-point scale: 0, poor quality and nonconfident; 1, fair quality and marginally confident; and 2, good quality and highly confident. The areas of stenosis or occlusion identified in each study were recorded on separate schematic diagrams for review. Ten vascular segments were evaluated, comprising the upper two thirds of the superficial femoral artery, the lower third of the superficial femoral artery and the above-knee popliteal artery, the below-knee popliteal artery, the upper third of the anterior tibial artery, the lower two thirds of the anterior tibial artery, the tibioperoneal trunk, the tibial posterior artery, the peroneal artery, the dorsal arteries of the foot, and the plantar arteries of the foot (the lateral and medial plantar arteries were interpreted together).

Segments were graded normal or stenosed less than 50% (group A), stenosed more than 50% (group B), or occluded (group C). For both techniques, the degree of stenosis was measured by dividing the minimal vessel luminal diameter in the segment by the maximal observed luminal diameter. Treatment plans formulated by each reviewer on the basis of MR angiography (i.e., bolus chase MR angiography and VIBE analysis) and DSA interpreted separately were further reported.

Statistical Analysis
Separate interpretations were used to calculate the sensitivity and specificity, with 95% confidence intervals (CIs), of our MR angiography protocol for detecting stenoses greater than 50% and arterial occlusions, using DSA as the gold standard. Intertechnique agreement was determined for the overall analysis, infrapopliteal analysis, and suprapopliteal analysis, by calculating Cohen's kappa coefficient. Interobserver agreement for each technique was determined by calculating Cohen's kappa coefficient for each observer. Kappa values of 0.81-1.00 correspond to near-perfect agreement.

Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Only arterial segments located below the common femoral artery were assessed, whatever the approach used. A total of 310 segments were finally assessed by each reviewer.

Analysis of Entire Vascular Tree
Compared with DSA, the sensitivity of hybrid MR angiography for depicting arterial stenosis greater than 50% (group B) ranged from 95% to 96%, and specificity was close to 98%. For arterial occlusion (group C), the sensitivity of hybrid MR angiography ranged from 90% to 95%, and specificity ranged from 98% to 99% (Table 2).

For both reviewers and for all locations, kappa values for intertechnique agreement were greater than 0.88, corresponding to near-perfect agreement (Table 3). Interobserver agreement was high for all locations and for both MR angiography and DSA. Kappa values for interobserver agreement on DSA were 0.97 for infrapopliteal segments, 1 for suprapopliteal vessels, and 0.98 overall. For MR angiography, kappa values for interobserver agreement were, respectively, 0.98, 0.98, and 0.98.

No differences in interpretation were found between MR angiography and DSA in 74% (23/31) of patients for reviewer 2, and in 71% (22/31) of patients for reviewer 1. Regarding the popliteal and suprapopliteal vessels, a perfect correlation between MR angiography and DSA was observed for 97% (30/31) of patients by the two reviewers.

Regarding segment-to-segment analysis, discrepant results were obtained in 31 (5%) of 620 comparisons. Of these 31 discrepancies, 28 (90%) involved infrapopliteal segments. Table 4 summarizes the results of segment-to-segment comparison of MR angiography and DSA.

Of the 31 discrepant results, 18 (58%) involved segments that were occluded (group C) on DSA but patent (group A) (n = 13) or stenosed (group B) (n = 5) on MR angiography. All these 18 segments were in infrapopliteal segments, 10 were distal to a long arterial occlusion, and four involved the dorsalis pedis or plantar artery (Figs. 1A, 1B, 1C, 1D, and 1E).

View larger version (92K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1A —48-year-old man with nonhealing ulceration of left calf. Anteroposterior bolus chase MR angiogram of calf shows marked venous enhancement (arrow) preventing adequate interpretation of distal arteries.

View larger version (56K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1B —48-year-old man with nonhealing ulceration of left calf. Anteroposterior volumetric interpolated breath-hold examination (VIBE) gadolinium-enhanced MR angiographic maximum-intensity-projection (MIP) image of left lower extremity shows patent but multistenosed anterior tibial artery (arrowheads). Dorsalis pedis artery appears to be patent and forms plantar arch. Venous enhancement (arrow) moderately affects interpretation based on this sole incidence. Lower two thirds of both peroneal artery and posterior tibial artery are not clearly identified.

View larger version (90K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1C —48-year-old man with nonhealing ulceration of left calf. Lateral VIBE gadolinium-enhanced MR angiographic MIP image confirms presence of left anterior tibial artery stenoses (arrowheads). Venous enhancement (arrow) is less present, allowing appropriate assessment of distal vessels. Lower two thirds of posterior tibial artery are patent and then are occluded above plantar arch. Peroneal artery is occluded at mid leg, and plantar artery is patent but has multiple stenoses.

View larger version (120K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1D —48-year-old man with nonhealing ulceration of left calf. Lower parts of anterior tibial artery and dorsalis pedis artery are not depicted on digital subtraction angiography (DSA), despite selective injection in left external iliac artery and delayed imaging. Correlation between MR angiography and DSA findings was graded as poor. G = left (gauche).

View larger version (134K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1E —48-year-old man with nonhealing ulceration of left calf. On the basis of MR angiography findings, percutaneous transluminal angioplasty (PTA) was performed. Lateral DSA of ankle and foot shortly after contrast injection in superficial femoral artery after PTA shows patency of both anterior tibial artery and dorsalis pedis artery (arrowheads). Posterior tibial artery remains occluded above ankle, and plantar arch does not fill.

View larger version (72K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 2A —60-year-old diabetic woman with focal gangrene of right calf. Lateral volumetric interpolated breath-hold examination (VIBE) gadolinium-enhanced MR angiographic maximum-intensity-projection image shows diffusely diseased anterior tibial artery, with stenosis of both its upper (arrowhead) and mid (arrow) portions, classified by both observers as greater than 50% (group B). Tibioperoneal trunk and entire length of posterior tibial artery are occluded. Distal anterior tibial artery is occluded, and peroneal artery fills into foot.

View larger version (126K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 2B —60-year-old diabetic woman with focal gangrene of right calf. Digital subtraction angiogram obtained after contrast injection in right external iliac artery confirms stenosis of upper anterior tibial artery (arrowhead) and occlusion of tibioperoneal trunk, but does not confirm findings regarding mid portion of anterior tibial artery (arrow), which is better depicted than on VIBE images. This segment was classified as belonging in group A by both observers. Marked calcifications were present along mid portion of anterior tibial artery.

View larger version (114K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3A —67-year-old diabetic man with nonhealing ulceration of left calf. Coronal 3D bolus chase MR angiograms of pelvis (A), thigh (B), and calf (C) compared with volumetric interpolated breath-hold examination (VIBE) (D) MR angiograms of left calf and ankle. Note venous contamination on MR angiograms of both left thigh and left calf (arrowheads, B and C). By comparison, no venous enhancement is seen on disease-free right calf (arrows, B and C). Because of early acquisition after bolus injection, venous overlay is not present on VIBE MR angiogram of left leg (D). Because of venous overlay, arterial vessels analysis was not possible on basis of bolus chase MR angiography findings but remained possible on VIBE angiography. Distal portion of anterior tibial artery is filled into foot, and pedal artery appears patent. Posterior tibial artery and distal portion of peroneal artery appear patent, but no plantar arch is filled.

View larger version (90K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3B —67-year-old diabetic man with nonhealing ulceration of left calf. Coronal 3D bolus chase MR angiograms of pelvis (A), thigh (B), and calf (C) compared with volumetric interpolated breath-hold examination (VIBE) (D) MR angiograms of left calf and ankle. Note venous contamination on MR angiograms of both left thigh and left calf (arrowheads, B and C). By comparison, no venous enhancement is seen on disease-free right calf (arrows, B and C). Because of early acquisition after bolus injection, venous overlay is not present on VIBE MR angiogram of left leg (D). Because of venous overlay, arterial vessels analysis was not possible on basis of bolus chase MR angiography findings but remained possible on VIBE angiography. Distal portion of anterior tibial artery is filled into foot, and pedal artery appears patent. Posterior tibial artery and distal portion of peroneal artery appear patent, but no plantar arch is filled.

View larger version (104K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3C —67-year-old diabetic man with nonhealing ulceration of left calf. Coronal 3D bolus chase MR angiograms of pelvis (A), thigh (B), and calf (C) compared with volumetric interpolated breath-hold examination (VIBE) (D) MR angiograms of left calf and ankle. Note venous contamination on MR angiograms of both left thigh and left calf (arrowheads, B and C). By comparison, no venous enhancement is seen on disease-free right calf (arrows, B and C). Because of early acquisition after bolus injection, venous overlay is not present on VIBE MR angiogram of left leg (D). Because of venous overlay, arterial vessels analysis was not possible on basis of bolus chase MR angiography findings but remained possible on VIBE angiography. Distal portion of anterior tibial artery is filled into foot, and pedal artery appears patent. Posterior tibial artery and distal portion of peroneal artery appear patent, but no plantar arch is filled.

View larger version (76K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3D —67-year-old diabetic man with nonhealing ulceration of left calf. Coronal 3D bolus chase MR angiograms of pelvis (A), thigh (B), and calf (C) compared with volumetric interpolated breath-hold examination (VIBE) (D) MR angiograms of left calf and ankle. Note venous contamination on MR angiograms of both left thigh and left calf (arrowheads, B and C). By comparison, no venous enhancement is seen on disease-free right calf (arrows, B and C). Because of early acquisition after bolus injection, venous overlay is not present on VIBE MR angiogram of left leg (D). Because of venous overlay, arterial vessels analysis was not possible on basis of bolus chase MR angiography findings but remained possible on VIBE angiography. Distal portion of anterior tibial artery is filled into foot, and pedal artery appears patent. Posterior tibial artery and distal portion of peroneal artery appear patent, but no plantar arch is filled.

View larger version (94K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4A —76-year-old diabetic women with focal gangrene of left calf. Coronal 3D bolus chase MR angiograms of thigh (A) and calf (B) compared with volumetric interpolated breath-hold examination (VIBE) MR angiogram of left calf and ankle (C) and selective digital subtraction angiograms (DSA) of thigh (D), calf (E), and ankle (F). Focal stenoses are seen in distal popliteal artery and tibioperoneal trunk (long arrows, A and C). Peroneal artery is patent into foot (short arrow, C). Proximal anterior tibial artery is diffusely diseased, reconstitutes above level of ankle, and becomes patent in foot. Note good correlation regarding short occlusion of distal left superficial femoral artery (arrowhead, A and D) depicted on both bolus-chase MR angiography (A) and selective DSA (D) and good correlation regarding popliteal artery stenoses depicted on both VIBE MR angiogram (arrows, C) and selective DSA (arrows, E). Left dorsalis pedis artery is depicted on VIBE MR angiograms (short arrow, C) and on selective DSA (short arrow, F; G = left [gauche]) but not on third-step bolus chase MR angiography (arrow, B). Note that because of calf pain, a slight change in patient's position between unenhanced and contrast-enhanced bolus chase MR angiography led to motion artifacts being clearly visible in soft tissues, diminishing accuracy of arterial vessel analysis (B).

View larger version (79K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4B —76-year-old diabetic women with focal gangrene of left calf. Coronal 3D bolus chase MR angiograms of thigh (A) and calf (B) compared with volumetric interpolated breath-hold examination (VIBE) MR angiogram of left calf and ankle (C) and selective digital subtraction angiograms (DSA) of thigh (D), calf (E), and ankle (F). Focal stenoses are seen in distal popliteal artery and tibioperoneal trunk (long arrows, A and C). Peroneal artery is patent into foot (short arrow, C). Proximal anterior tibial artery is diffusely diseased, reconstitutes above level of ankle, and becomes patent in foot. Note good correlation regarding short occlusion of distal left superficial femoral artery (arrowhead, A and D) depicted on both bolus-chase MR angiography (A) and selective DSA (D) and good correlation regarding popliteal artery stenoses depicted on both VIBE MR angiogram (arrows, C) and selective DSA (arrows, E). Left dorsalis pedis artery is depicted on VIBE MR angiograms (short arrow, C) and on selective DSA (short arrow, F; G = left [gauche]) but not on third-step bolus chase MR angiography (arrow, B). Note that because of calf pain, a slight change in patient's position between unenhanced and contrast-enhanced bolus chase MR angiography led to motion artifacts being clearly visible in soft tissues, diminishing accuracy of arterial vessel analysis (B).

View larger version (92K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4C —76-year-old diabetic women with focal gangrene of left calf. Coronal 3D bolus chase MR angiograms of thigh (A) and calf (B) compared with volumetric interpolated breath-hold examination (VIBE) MR angiogram of left calf and ankle (C) and selective digital subtraction angiograms (DSA) of thigh (D), calf (E), and ankle (F). Focal stenoses are seen in distal popliteal artery and tibioperoneal trunk (long arrows, A and C). Peroneal artery is patent into foot (short arrow, C). Proximal anterior tibial artery is diffusely diseased, reconstitutes above level of ankle, and becomes patent in foot. Note good correlation regarding short occlusion of distal left superficial femoral artery (arrowhead, A and D) depicted on both bolus-chase MR angiography (A) and selective DSA (D) and good correlation regarding popliteal artery stenoses depicted on both VIBE MR angiogram (arrows, C) and selective DSA (arrows, E). Left dorsalis pedis artery is depicted on VIBE MR angiograms (short arrow, C) and on selective DSA (short arrow, F; G = left [gauche]) but not on third-step bolus chase MR angiography (arrow, B). Note that because of calf pain, a slight change in patient's position between unenhanced and contrast-enhanced bolus chase MR angiography led to motion artifacts being clearly visible in soft tissues, diminishing accuracy of arterial vessel analysis (B).

View larger version (91K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4D —76-year-old diabetic women with focal gangrene of left calf. Coronal 3D bolus chase MR angiograms of thigh (A) and calf (B) compared with volumetric interpolated breath-hold examination (VIBE) MR angiogram of left calf and ankle (C) and selective digital subtraction angiograms (DSA) of thigh (D), calf (E), and ankle (F). Focal stenoses are seen in distal popliteal artery and tibioperoneal trunk (long arrows, A and C). Peroneal artery is patent into foot (short arrow, C). Proximal anterior tibial artery is diffusely diseased, reconstitutes above level of ankle, and becomes patent in foot. Note good correlation regarding short occlusion of distal left superficial femoral artery (arrowhead, A and D) depicted on both bolus-chase MR angiography (A) and selective DSA (D) and good correlation regarding popliteal artery stenoses depicted on both VIBE MR angiogram (arrows, C) and selective DSA (arrows, E). Left dorsalis pedis artery is depicted on VIBE MR angiograms (short arrow, C) and on selective DSA (short arrow, F; G = left [gauche]) but not on third-step bolus chase MR angiography (arrow, B). Note that because of calf pain, a slight change in patient's position between unenhanced and contrast-enhanced bolus chase MR angiography led to motion artifacts being clearly visible in soft tissues, diminishing accuracy of arterial vessel analysis (B).

View larger version (79K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4E —76-year-old diabetic women with focal gangrene of left calf. Coronal 3D bolus chase MR angiograms of thigh (A) and calf (B) compared with volumetric interpolated breath-hold examination (VIBE) MR angiogram of left calf and ankle (C) and selective digital subtraction angiograms (DSA) of thigh (D), calf (E), and ankle (F). Focal stenoses are seen in distal popliteal artery and tibioperoneal trunk (long arrows, A and C). Peroneal artery is patent into foot (short arrow, C). Proximal anterior tibial artery is diffusely diseased, reconstitutes above level of ankle, and becomes patent in foot. Note good correlation regarding short occlusion of distal left superficial femoral artery (arrowhead, A and D) depicted on both bolus-chase MR angiography (A) and selective DSA (D) and good correlation regarding popliteal artery stenoses depicted on both VIBE MR angiogram (arrows, C) and selective DSA (arrows, E). Left dorsalis pedis artery is depicted on VIBE MR angiograms (short arrow, C) and on selective DSA (short arrow, F; G = left [gauche]) but not on third-step bolus chase MR angiography (arrow, B). Note that because of calf pain, a slight change in patient's position between unenhanced and contrast-enhanced bolus chase MR angiography led to motion artifacts being clearly visible in soft tissues, diminishing accuracy of arterial vessel analysis (B).

View larger version (99K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4F —76-year-old diabetic women with focal gangrene of left calf. Coronal 3D bolus chase MR angiograms of thigh (A) and calf (B) compared with volumetric interpolated breath-hold examination (VIBE) MR angiogram of left calf and ankle (C) and selective digital subtraction angiograms (DSA) of thigh (D), calf (E), and ankle (F). Focal stenoses are seen in distal popliteal artery and tibioperoneal trunk (long arrows, A and C). Peroneal artery is patent into foot (short arrow, C). Proximal anterior tibial artery is diffusely diseased, reconstitutes above level of ankle, and becomes patent in foot. Note good correlation regarding short occlusion of distal left superficial femoral artery (arrowhead, A and D) depicted on both bolus-chase MR angiography (A) and selective DSA (D) and good correlation regarding popliteal artery stenoses depicted on both VIBE MR angiogram (arrows, C) and selective DSA (arrows, E). Left dorsalis pedis artery is depicted on VIBE MR angiograms (short arrow, C) and on selective DSA (short arrow, F; G = left [gauche]) but not on third-step bolus chase MR angiography (arrow, B). Note that because of calf pain, a slight change in patient's position between unenhanced and contrast-enhanced bolus chase MR angiography led to motion artifacts being clearly visible in soft tissues, diminishing accuracy of arterial vessel analysis (B).

Therapeutic Implications
The two reviewers planned a total of 83 treatments (reviewer 1, 41; reviewer 2, 42) on the basis of DSA, and 94 treatments (reviewer 1, 46; reviewer 2, 48) on the basis of MR angiography (several treatments could be proposed for the same patient). Of these, 80% (66/83) and 79% (74/94) concerned calf and dorsalis pedis arteries on DSA and MR angiography, respectively. All treatments proposed on the basis of DSA findings were endorsed by MR angiography findings.

Eleven discordant treatments (six patients) were proposed by the two reviewers: seven cases (four patients) concerned below-the-knee segments that were graded group C on DSA but group A or B on MR angiography. In four of these seven cases (two patients), the dorsalis pedis or plantar artery was detected on MR angiography, potentially indicating the possibility of distal bypass or endovascular revascularization. Four of the 11 discrepant cases (two patients) involved interpretation of group A stenoses (in two superficial femoral arteries and two anterior tibial arteries), all of which were interpreted as being more severe on MR angiography (Figs. 2A, and 2B). In all these cases, marked calcifications of the stenosis were observed.

Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
These results suggest that the diagnostic accuracy of MR angiography in the assessment of critical limb ischemia in selected diabetic patients can be improved by using a hybrid protocol. MR angiography has markedly changed the diagnostic approach to arterial disease since the advent of contrast-enhanced ultrafast studies in 1995 [22]. Recent studies suggest the superiority of contrast-enhanced MR angiography over DSA for the identification of patent arterial segments in runoff vessels of the foot in both diabetic and nondiabetic patients [11, 14, 15, 23]. Dorweiler et al. [12] showed that foot vessels that were not seen on conventional angiography but were detected on MR angiography were suitable target vessels for durable pedal bypass grafting. The superiority of MR angiography is particularly clear in infrapopliteal vessels [12, 23]. Several explanations have been proposed for this shortcoming of DSA: MR angiography can image blood flow at velocities as low as 2 cm/sec, whereas in DSA the dilution of contrast medium below long-occluded segments leads to insufficient enhancement of distal vessels [21].

No single technique for peripheral MR angiography has been universally accepted. Some authors have reported the feasibility of peripheral MR angiography with multiple stacks. However, this technique has not gained widespread clinical acceptance for various reasons: Imaging the entire vascular tree requires both repeated placement of the patient in the scanner and repeated injections of contrast media. This results in longer examination times and a lower contrast-to-noise ratio after the first stack because of increased contrast enhancement of surrounding tissue. Furthermore, the accuracy of arterial analysis is compromised by previous venous overlay.

The introduction of the three-station technique, using a moving bed, theoretically eliminates these shortcomings. For the two proximal arterial regions, sensitivity and specificity exceed 90% with bolus chase techniques [24, 25]. However, the third step (calf and dorsalis pedis artery assessment) is performed approximately 40 sec after opacification of the aorta, and acquisition lasts 20-30 sec. This is too long in critical limb ischemia associated with cellulitis and ulceration, particularly in patients with diabetes.

Prince et al. [17] have shown that patients with type 2 diabetes and cellulitis or ulcerations tend to have faster arterial flow. This faster flow in patients with tissue loss associated with diabetes has been tentatively explained in several ways: vascular calcifications can result in decreased arterial compliance, and cellulitis or ulceration of the lower extremity may result in vasodilatation and increased demand for flow [17]. The result is early venous enhancement, which decreases the accuracy of MR angiography, particularly for the characterization of calf and ankle arteries [18, 19].

A hybrid protocol, focusing separately on pelvis-to-thigh arteries and calf and dorsalis pedis arteries, was recently described [15, 26] with promising results. Although there was no significant difference in diagnostic quality of the pelvis and thigh arterial segments between the bolus chase and hybrid techniques, the diagnostic quality of calf images was significantly higher with the hybrid technique. This improvement in quality correlated with significantly higher sensitivity and specificity of calf analysis (100% and 91%, respectively) with the hybrid technique compared with the bolus chase technique alone [15].

However, in the study of Morasch et al. [15], only 58% of patients had critical ischemia, and no information on diabetes was provided. Recently, time-resolved acquisitions such as TRICKS (time-resolved imaging of contrast kinetics) sequences have been developed [27]. The high temporal resolution thus obtained permits pure arterial phase imaging [27]. However, these techniques need to be evaluated in diabetic patients.

Our study confirms the good performance of a hybrid MR angiography protocol in peripheral artery disease associated with critical limb ischemia in diabetic patients. The use of a body surface coil for limb coverage ensured adequate visualization of dorsalis pedis arteries. Furthermore, maximal external hip rotation and moderate knee flexion allowed both the limb and the foot to remain parallel to the phased-array spine coil. This position also permits plantar flexion of the feet to be maintained, which seems to give better visualization of pedal vessels [15].

Fat saturation was used, first, to increase the contrast between enhanced vessels and surrounding tissues and, second, to reduce reliance on subtraction for the elimination of high signal intensities from subcutaneous fat and bone marrow [18]. Our first-step VIBE analysis of calf and dorsalis pedis arteries thus yielded images with good quality, even when bolus chase MR angiography failed to depict runoff vessels. Our results confirm that an optimized hybrid MR angiography protocol can reveal dorsalis pedis or plantar arteries possibly amenable to treatment. Hybrid MR angiography enables an initial acquisition focused on distal calf vessels, thus reducing the presence of early venous enhancement more frequently encountered with bolus chase MR angiography. Furthermore, optimal detection of contrast medium in distal vessels is achieved through the CARE bolus technique applied to the hybrid MR angiography protocol.

Recently, MDCT angiography was proposed for the evaluation of peripheral vascular disease [29]. New developments in multidectector technology increase the speed of acquisition and offer high spatial resolution. MDCT of the entire arterial supply of the legs, from the suprarenal aorta to the ankles in a single helical acquisition, is now possible. However, misinterpretations are observed in cases of severe diffuse calcification, particularly in small vessels such as calf or ankle vessels [28]. Moreover, MDCT angiography requires injection of iodinated contrast material, with the risk of contrast-related renal failure.

Several limitations may have affected the results and conclusions of our study. To perform the therapeutic procedure, the DSA operators were aware of the MR angiography results. However, we systematically used a standardized protocol for DSA and postprocessing, including selection of a predetermined volume and rate of contrast injection, field of view, and projection incidences. Because no additional series or projection was allowed, the performance of DSA could have been suboptimal. However, such DSA protocols are routinely used for patients with diabetes and renal impairments [11, 14, 27].

Because of its lack of nephrotoxicity, CO₂ has been proposed as an alternative to iodine-based contrast agents in peripheral angiography in diabetic patients with compromised renal function. CO₂ is less viscous than iodine-based contrast material and does not dilute, resulting in a better filling of distal vessels in the case of proximal occlusion. However, in our study, assessment of distal arteries was not impaired by proximal artery occlusion and thus did not require the use of CO₂ DSA. Furthermore, using CO₂ DSA only in selected patients with proximal artery occlusion would have biased our study by changing the standardized DSA procedure. None of the discrepant cases in which calf and dorsalis pedis arteries were not detected on DSA but were patent on MR angiography concerned DSA procedures performed on the OEC unit.

We acknowledge that DSA results might have been improved by the injection of contrast medium in the distal vessels such as the distal superficial femoral artery or popliteal artery and by performing multiple acquisitions with different projection views. However, our aim was to compare standardized procedures, that is, standard DSA with bolus chase MR angiography and hybrid MR angiography. Our study thus highlights that both standard DSA and standard bolus chase MR angiography are insufficient for adequate assessment of distal runoff vessels in diabetic patients with critical limb ischemia. Furthermore, our study suggests that to fully explore such distal vessels, DSA would probably require additional contrast medium injection at the expense of potentially higher renal toxicity.

Because our hybrid MR angiography protocol requires a 60- to 90-min interruption, its practicability could be challenged. But assessment of calf and foot arteries was inconclusive on bolus chase MR angiography alone in 50 (81%) of 62 analyses. We believe that these findings alone justify an additional waiting time for optimal diagnostic performance. Furthermore, the interruption could probably be shortened, depending on the clearance of the contrast agent from distal veins [28].

Another potential limitation for the widespread clinical use of our hybrid MR angiography protocol is the high rate of contraindications in our study population. We deliberately excluded patients with intraarterial stents, which is not a contraindication to MR angiography. Signal loss in the stent impairs the analysis of the vessel lumen [29, 30], which could have been considered a potential bias of this study.

Consensus analysis between the two reviewers was not used; therefore, our results reflect both the performance of this technique and the interpretation of the images that the technique produced. With our protocol, interpretation was reliable whether the observer was a radiologist or a vascular surgeon. The value of Cohen's kappa coefficient was always consistent, with almost perfect intertechnique agreement.

During DSA, we did not use a vasodilating agent because we speculated that the calf vessels would already be maximally dilated in chronic limb ischemia associated with diabetes. In addition, there is no convincing evidence that vasodilating drugs improve image quality [31, 32].

Recently, new improvements of MR angiography have been described. Partially parallel acquisition imaging offers faster acquisition and will probably find many applications in MR angiography [33]. With the application of improvements in high gradient technology and the use of dedicated vascular contrast medium, the accuracy of MR angiography will probably improve.

In conclusion, our results suggest that hybrid MR angiography is a reliable method for investigating peripheral artery disease in selected diabetic patients with critical limb ischemia. Moreover, hybrid MR angiography visualizes lower extremity vessels that are not seen on conventional angiography, avoids iodinated contrast material-induced renal failure, and may be useful for treatment planning in this setting.

References

Top Abstract Introduction Subjects and Methods Results Discussion References

 

1. Dyett JF, Nicholson AA, Ettles DF. Vascular imaging and intervention in peripheral arteries in the diabetic patient. Diabetes Metab Res Rev 2000;16 [suppl 1]:S16 -S22
2. Rubba P, Leccia G, Faccenda F, et al. Diabetes mellitus and localizations of obliterating arterial disease of the lower limbs. Angiology 1991;42 : 296-301
3. van der Feen C, Neijens FS, Kanters SD, Mali WP, Stolk RP, Banga JD. Angiographic distribution of lower extremity atherosclerosis patients with and without diabetes. Diabet Med 2002;19 : 366-370[Medline]
4. Jude EB, Oyibo SO, Chalmers N, Boulton AJ. Peripheral arterial disease in diabetic and nondiabetic patients: a comparison of severity and outcome. Diabetes Care 2001;24 : 1433-1437[Abstract/Free Full Text]
5. LoGerfo FW, Coffman JD. Vascular and microvascular disease of the foot in diabetes: implications for foot care. N Engl J Med 1984; 311:1615 -1619[Medline]
6. Hanna GP, Fujise K, Kjellgren O, et al. Infrapopliteal transcatheter interventions for limb salvage in diabetic patients: importance of aggressive interventional approach and role of transcutaneous oximetry. J Am Coll Cardiol 1997;30 : 664-669[Abstract]
7. Pomposelli FB, Marcaccio EJ, Gibbons GW, et al. Dorsalis pedis arterial bypass: durable limb salvage for foot ischemia in patients with diabetes mellitus. J Vasc Surg 1995;21 : 375-384[CrossRef][Medline]
8. Melliere D, Berrahal D, Desgranges P, et al. Influence of diabetes on revascularization procedures of the aorta and lower limb arteries: early results. Eur J Vasc Endovasc Surg 1999;17 : 438-441[Medline]
9. Faglia E, Favales F, Quarantiello A, et al. Angiographic evaluation of peripheral arterial occlusive disease and its role as a prognostic determinant for major amputation in diabetic subjects with foot ulcers. Diabetes Care 1998;21 : 625-630[Abstract]
10. Toursarkissian B, D'Ayala M, Stefanidis D, et al. Angiographic scoring of vascular occlusive disease in the diabetic foot: relevance to bypass graft patency and limb salvage. J Vasc Surg2002; 35:494 -500[Medline]
11. Kreitner KF, Kalden P, Neufang A, et al. Diabetes and peripheral arterial occlusive disease: prospective comparison of contrast-enhanced three-dimensional MR angiography with conventional digital subtraction angiography. AJR 2000;174 : 171-179[Abstract/Free Full Text]
12. Dorweiler B, Neufang A, Kreitner KF, Schmiedt W, Oelert H. Magnetic resonance angiography unmasks reliable target vessels for pedal bypass grafting in patients with diabetes mellitus. J Vasc Surg 2002; 35:766 -772[CrossRef][Medline]
13. Koelemay MJ, Lijmer JG, Stoker J, Legemate DA, Bossuyt PM. Magnetic resonance angiography for the evaluation of lower extremity arterial disease: a meta-analysis. JAMA 2001;285 : 1338-1345[Abstract/Free Full Text]
14. Loewe C, Schoder M, Rand T, et al. Peripheral vascular occlusive disease: evaluation with contrast-enhanced moving-bed MR angiography versus digital subtraction angiography in 106 patients. AJR2002; 179:1013 -1021[Abstract/Free Full Text]
15. Morasch MD, Collins J, Pereles FS, et al. Lower extremity stepping-table magnetic resonance angiography with multilevel contrast timing and segmented contrast infusion. J Vasc Surg2003; 37:62 -71[CrossRef][Medline]
16. Wang Y, Winchester PA, Khilnani NM, et al. Contrast-enhanced peripheral MR angiography from abdominal aorta to pedal arteries: combined dynamic two-dimensional acquisitions and bolus-chase three-dimensional acquisitions. Invest Radiol 2001;36 : 170-177[CrossRef][Medline]
17. Prince MR, Chabra SG, Watts R, et al. Contrast material travel times in patients undergoing peripheral MR angiography. Radiology 2002;224 : 55-61[Abstract/Free Full Text]
18. Foo TK, Ho VB, Hood MN, Marcos HB, Hess SL, Choyke PL. High-spatial-resolution multistation MR imaging of lower-extremity peripheral vasculature with segmented volume acquisition: feasibility study. Radiology 2001;219 : 835-841[Abstract/Free Full Text]
19. Hany TF, Carroll TJ, Omary RA, et al. Aorta and runoff vessels: single-injection MR angiography with automated table movement compared with multiinjection time-resolved MR angiography—initial results. Radiology 2001;221 : 266-272[Abstract/Free Full Text]
20. Rutherford RB, Baker JD, Ernst C, et al. Recommended standards for reports dealing with lower extremity ischemia: revised version. J Vasc Surg 1997; 26:517 -538[CrossRef][Medline]
21. Gates J, Hartnell GG. Optimized diagnostic angiography in high-risk patients with severe peripheral vascular disease. RadioGraphics 2000;20 : 121-133[Abstract/Free Full Text]
22. Prince MR, Narasimham DL, Stanley JC, et al. Breath-hold gadolinium-enhanced MR angiography of the aorta and its major branches. Radiology 1995;197 : 785-792[Abstract/Free Full Text]
23. Owen RS, Carpenter JP, Baum RA, Perloff LJ, Cope C. Magnetic resonance imaging of angiographically occult runoff vessels in peripheral arterial occlusive disease. N Engl J Med1992; 326:1577 -1581[Abstract]
24. Ho KY, Leiner T, de Haan MW, Kessels AG, Kitslaar PJ, van Engelshoven JM. Peripheral vascular tree stenoses: evaluation with moving-bed infusion-tracking MR angiography. Radiology1998; 206:683 -692[Abstract/Free Full Text]
25. Meaney JF, Ridgway JP, Chakraverty S, et al. Stepping-table gadolinium-enhanced digital subtraction MR angiography of the aorta and lower extremity arteries: preliminary experience. Radiology1999; 211:59 -67[Abstract/Free Full Text]
26. Binkert CA, Baker PD, Petersen BD, Szumowski J, Kaufman JA. Peripheral vascular disease: blinded study of dedicated calf MR angiography versus standard bolus-chase MR angiography and film hard-copy angiography. Radiology 2004;232 : 860-866[Abstract/Free Full Text]
27. Swan JS, Carroll TJ, Kennell TW, et al. Time-resolved three-dimensional contrast-enhanced MR angiography of the peripheral vessels. Radiology 2002;225 : 43-52[Abstract/Free Full Text]
28. Martin ML, Tay KH, Flak B, et al. Multidetector CT angiography of the aortoiliac system and lower extremities: a prospective comparison with digital subtraction angiography. AJR2003; 180:1085 -1091[Abstract/Free Full Text]
29. Wang Y, Truong TN, Yen C, et al. Quantitative evaluation of susceptibility and shielding effects of nitinol, platinum, cobalt-alloy, and stainless-steel stents. Magn Reson Med2003; 49:972 -976[CrossRef][Medline]
30. Bartels LW, Smits H, Bakker CJ, Viergever MA. MR imaging of vascular stents: effects of susceptibility, flow, and radiofrequency eddy currents. J Vasc Interv Radiol 2001;12 : 365-371[Medline]
31. London NJ, Varty K, Sayers RD, Thompson MM, Bell PR, Bolia A. Percutaneous transluminal angioplasty for lower-limb critical ischaemia. Br J Surg 1995;82 : 1232-1235[Medline]
32. Alson MD, Lang EV, Kaufman JA. Pedal arterial imaging. J Vasc Interv Radiol 1997; 8:9 -18[Medline]
33. Sodickson DK, Manning WJ. Simultaneous acquisition of spatial harmonics (SMASH): fast imaging with radiofrequency coil arrays. Magn Reson Med 1997;38 : 591-603[Medline]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

H. Ersoy and F. J. Rybicki MR Angiography of the Lower Extremities Am. J. Roentgenol., June 1, 2008; 190(6): 1675 - 1684. [Abstract]