| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cardiac Imaging |
1 From the Departments of Diagnostic Radiology (M.F., N.I.S., U.K., C.D.C., S.M.) and Internal Medicine, Division of Cardiology (A.F., U.H.), Eberhard-Karls-University Tuebingen, Hoppe-Seyler-Str 3, 72076 Tuebingen, Germany. Received June 9, 2004; revision requested August 20; revision received October 20; accepted December 10. Address correspondence to M.F. (e-mail: michael.fenchel{at}med.uni-tuebingen.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Local ethics committee approval and informed consent were obtained. Rest-stress perfusion MR imaging studies were performed in 18 patients with coronary artery disease (12 men, six women; mean age, 58.6 years ± 13.6 [standard deviation]; range, 30–79 years) at 1.5 T with a multisection saturation-recovery SSFP sequence and 0.025 mmol gadopentetate dimeglumine per kilogram of body weight. MR studies were performed before (n = 18), several days after (n = 18), and 8 months after (n = 10) coronary intervention. Nine patients underwent percutaneous transluminal coronary angioplasty (PTCA) alone, and nine patients underwent PTCA with stent placement. Myocardial perfusion reserve index (MPRI) was calculated by dividing results of myocardial perfusion at maximal vasodilation by results at rest. The standard for myocardial perfusion was technetium 99m tetrofosmin single photon emission computed tomography. Statistical significance was tested with univariate variance analysis and Student t tests.
RESULTS: In the area of the stenosed vessel, MPRI was 1.04 ± 0.24 before treatment and 2.18 ± 0.57 several days afterward (P < .001). In remote areas, MPRI was 2.42 ± 0.44. In the stent group, MPRI increased by 156%, from 0.99 ± 0.20 before stent placement to 2.53 ± 0.53 after (P < .001). Similarly, in the PTCA only group, MPRI increased by 72%, from 1.08 ± 0.27 before PTCA to 1.87 ± 0.39 after (P < .001). At follow-up in patients without recurring chest pain, MPRI was 2.14 ± 0.37 in the area of the treated artery and 2.29 ± 0.47 in remote areas (P = .06).
CONCLUSION: The MPRI, derived from rest-stress examinations, can provide information on success of interventional procedures in stenosed coronary arteries.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Results of several studies have shown that magnetic resonance (MR) imaging is a valuable tool in the assessment of myocardial perfusion (13–19). Despite the potential important effect of perfusion MR imaging on patient care, to our knowledge only a few studies have been performed to investigate the ability to identify changes in myocardial perfusion over time. However, this ability would be of clinical value in patient assessment after conventional revascularization procedures such as PTCA or coronary bypass surgery. Taking into account that reduction of myocardial perfusion occurs earlier in the ischemic cascade than do left ventricular wall motion abnormalities, MR perfusion imaging appears to be even more promising (20). Considering the limitations of scintigraphic methods, such as radiation dose or time requirements, a noninvasive follow-up of patients may help to decide whether conventional coronary angiography is required.
Authors of recent studies have reported improved image quality and signal-to-noise ratio with steady-state free precession (SSFP) perfusion MR sequences in comparison with spoiled gradient-echo perfusion sequences (21–24). Thus, the purpose of our study was to prospectively assess changes in myocardial perfusion before and after successful coronary artery intervention in patients with single-vessel coronary artery disease by using an SSFP perfusion MR imaging sequence.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Findings at coronary angiography revealed single-vessel disease in 15 patients (left anterior descending artery, n = 14; left circumflex artery, n = 1) and two-vessel disease in three patients (both left anterior descending and left circumflex). Eleven patients had a history of prior myocardial infarction (anterior wall, n = 10; posterior wall, n = 1).
In all patients, MR perfusion examinations were performed immediately before intervention and several days after successful balloon dilation or stent placement (mean interval, 10.8 days ± 6.1; range, 3–22 days). Nine patients were treated by means of PTCA alone (hereafter, PTCA only group), whereas the other nine patients were treated by means of a stent placed into the lumen of the stenosed coronary artery after PTCA (hereafter, stent group). Coronary intervention was performed with or without stent placement on the basis of the decision of the physician performing the procedure.
In 10 patients, follow-up MR perfusion imaging was performed approximately 8 months (8.1 months ± 1.9) after coronary intervention, whereas the other eight patients refused the follow-up MR examination. Five patients from the stent group and five patients from the PTCA only group participated in the follow-up examinations.
Coronary Angiography and Image Interpretation
Coronary angiography was performed with a conventional angiography unit (Integris H; Philips Medical Systems, Best, the Netherlands) by using 5-F high-flow Judkins or Amplatz catheters (Cordis, Miami Lakes, Fla) and acquiring images in multiple projections. Coronary artery stenoses were imaged in the center of the field from multiple projections, and overlap of side branches and foreshortening of relevant coronary arteries was avoided as much as possible. After selecting the projection that showed maximal severity, the luminal diameter of the stenosed artery, along with adjacent reference segments, was measured on the end-diastolic frame. The severity of the stenosis was expressed as a percentage reduction of the internal diameter in relation to the estimated diameter interpolated from the diameters at the proximal and distal boundaries of the stenosis. A reduction of 75% or more in the luminal diameter was considered significant. Area was calculated as follows: area = 
(diameter/2)2. The area stenosis was calculated as follows: area stenosis = areaNML – areaRES, where areaNML and areaRES indicate the areas of the normal lumen and the residual lumen, respectively.
The analysis of the angiograms was performed by an experienced cardiologist (U.H., 8 years of experience) who was blinded to patient history and the results of the MR examination.
MR Imaging Protocol
MR imaging was performed by using a 1.5-T system (Magnetom Sonata; Siemens Medical Systems, Erlangen, Germany) equipped with high-performance gradients (maximum amplitude, 40 mT/m; slew rate, 200 mT/[m · msec]). An 18-gauge catheter was inserted into an antecubital vein for injection of the contrast agent. MR-compatible electrocardiography leads were placed on the patient's chest. Imaging was performed by using a phased-array surface coil as a receiver.
After localizing images were obtained, breath-hold SSFP cine images were acquired in multiple orientations, including the vertical and horizontal long axis and the short axis. Sequence parameters were as follows: 3.08/1.54 (repetition time msec/echo time msec), flip angle of 46°, section thickness of 5 mm, and temporal resolution of 46 msec.
Consecutive short-axis sections were acquired from the base of the heart to the apex in increments of 10 mm. Cine images were used to diagnose myocardial wall motion abnormalities in patients with coronary artery disease.
Left ventricular short-axis orientation was used for perfusion MR imaging (to minimize partial volume effects). For myocardial perfusion measurement, three representative short-axis sections were obtained, one each in the basal, midventricular, and apical regions of the left ventricle. In some patients, because of a fast heart rate, only two short-axis sections could be measured (midventricular and apical regions of the left ventricle). Acquisition of perfusion MR images was started simultaneously with injection of 0.025 mmol gadopentetate dimeglumine per kilogram of body weight (Magnevist; Schering, Berlin, Germany); perfusion images were obtained by using a two-dimensional saturation-recovery SSFP sequence. The bolus injection of gadopentetate dimeglumine (at a flow rate of 5 mL/sec) was followed with a 20-mL flush of 0.9% NaCl (at a flow rate of 5 mL/sec). Perfusion series images were acquired with a temporal resolution of one image per section per heartbeat. A delay of 15 minutes after the rest examination allowed residual gadopentetate dimeglumine to be washed out of the myocardium. Pharmacologic stress was applied by using dipyridamole (Persantine; Boehringer Ingelheim, Ingelheim, Germany) according to a standardized protocol (0.56 mg dipyridamole per kilogram of body weight over 4 minutes). Images were obtained with the same orientation and position before and after the administration of dipyridamole. Furthermore, the stress examination was performed by using identical measurement and injection parameters as before. Because dipyridamole leads to a reflectory tachycardia, an increase of 20% or more in the patient's heart rate served as control for a sufficient pharmacologic stimulation of perfusion reserve.
Subsequently, 0.1 mmol gadopentetate dimeglumine per kilogram of body weight was administered to visualize infarcted areas. After 15 minutes, delayed enhancement images were acquired in several short-axis sections by using an inversion-recovery gradient-recalled echo MR sequence. Parameters for inversion-recovery gradient-recalled echo sequences were as follows: 9.56/4.38/200–260 (repetition time msec/echo time msec/inversion time msec); field of view, 300–340 mm; flip angle, 25°; matrix, 166 x 256; and section thickness, 5 mm. The inversion time was chosen to minimize the signal from normal myocardium.
Perfusion Sequence
The two-dimensional saturation-recovery SSFP perfusion MR sequence was prospectively electrocardiographically triggered. The sequence consisted of a non–section-selective 90° saturation pulse and single-shot SSFP image acquisition with 2.4/1.2, flip angle of 50°–60° (depending on specific absorption rate limits), bandwidth of 1300 Hz per pixel, matrix of 72 x 128, and field of view of 225 x 300 mm. This rendered an inplane resolution of 1.55 x 1.15 mm after interpolation with zero filling for all images. Section thickness was 8 mm.
Technetium 99m Tetrofosmin SPECT and Interpretation
Single photon emission computed tomography (SPECT) studies were performed within 2 weeks prior to the preinterventional MR perfusion study. A standardized 1-day protocol was used. Pharmacologic stress was applied by means of injection of 0.56 mg dipyridamole per kilogram of body weight over 4 minutes. A bolus of 800 MBq technetium 99m (99mTc) tetrofosmin was injected in an antecubital vein and was flushed with 10 mL of 0.9% NaCl. SPECT image acquisition was performed with the patient in the supine position by using a double-headed rotating gamma camera (Millenium VG; GE Medical Systems, Waukesha, Wis) with a high-resolution collimator. All images were reconstructed in short-axis and in vertical and horizontal long-axis views. For data evaluation, short-axis sections corresponding to perfusion MR sections were used.
Two experienced observers (one with 3 years of experience in diagnostic radiology [U.K.], one with 4 years of experience in nuclear medicine) who were blinded to patient history and results from other imaging modalities evaluated the SPECT images. SPECT sections corresponding to sections from perfusion MR studies were divided into 12 radial segments, and each segment was rated as hypoperfused ("SPECT hypo") or normal ("SPECT norm") at rest and stress studies.
MR Data Analysis
Matlab software (MathWorks, Natick, Mass) was used to analyze myocardial perfusion (M.F., 3 years of experience with cardiac MR imaging). The myocardium of short-axis sections was subdivided into 12 radial segments, and signal intensity–time curves were calculated separately for each segment. Baseline signal intensity (SIBA) measured before contrast material administration was used to correct for any coil-induced differences in signal intensity in the segments: SINML = (SICON – SIBA)/SIBA, where SINML and SICON are the normalized signal intensity and the signal intensity after contrast agent administration, respectively. A least-square curve fit was performed for each raw data signal intensity–time curve by using a cubic polynomial function. The fitting procedure was observed for each segment (raw data points together with fitted curve), and the "goodness of fit" was assessed in calculating the residual sum of squared differences between the fitted curve and the data points, which had to be below 5% of the sum of signal intensities.
By using this 12-segment model, two segments located in the center of the area supplied by the stenosed coronary artery ("SPECT hypo") and two segments from the opposite or remote part of the left ventricle ("SPECT norm") were used for the evaluation process. Care was taken that the segments were not within a region showing delayed enhancement. For the semiquantitative evaluation, the parameter of maximum upslope was used from the fitted signal intensity–time curves. A myocardial perfusion reserve index (MPRI) was calculated from rest and stress perfusion data for each segment by dividing upslope during stress by upslope at rest.
Statistical Analysis
If not stated otherwise, data are given as mean ± standard deviation. P 
.05 was considered to indicate a statistically significant difference.
An estimation of the necessary number of patients in each group was performed before the start of the study by using differences in myocardial perfusion found in other patients at imaging examinations performed before and after intervention. This yielded a total sample size of 10 patients as sufficient for this study. Furthermore, by using the data from the present study and assuming 
= .05 and ß = .20 (power of 80%), the result was a sample size of nine patients each for the PTCA only group and the stent group.
All data used for statistical calculations were tested for normal distribution by using a Kolmogorov-Smirnov test. A univariate variance analysis was performed to detect statistically significant effects on perfusion values between different time points (before treatment, after treatment, and at follow-up), regions (stenosed and remote areas), and procedures (PTCA only or PTCA with stent placement) by using SPSS software (11.0; SPSS, Chicago, Ill). Differences in MPRI values from stenosed and remote areas were tested for significance by using a paired Student t test. Differences between the stent group and the PTCA only group were tested by using an unpaired Student t test. Furthermore, MPRI values in segments related to the treated coronary artery were tested with a paired Student t test against the MPRI values from the same segments before intervention. Likewise, MPRI values from control segments before and after coronary intervention were compared. The Student t tests were performed by using the software JMP (version 4; SAS Institute, Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
After administration of dipyridamole, patients' heart rates increased significantly: Before coronary intervention, heart rate increased from a mean of 63.8 beats per minute ± 12.0 to 82.6 beats per minute ± 13.1 after administration of dipyridamole, and after coronary intervention, heart rate increased from 62.5 beats per minute ± 8.6 to 78.4 beats per minute ± 9.9 after administration of dipyridamole.
Conventional Angiography and Intervention
Coronary angiography of the diseased arteries revealed a mean luminal stenosis of 89% ± 13 and a mean diameter of 0.89 mm ± 0.51 before intervention (Table), as well as a mean diameter of 3.74 mm ± 1.37 after intervention. Concerning area stenosis and preprocedural minimal luminal diameter, no significant difference was found between patients in the PTCA only group and those in the stent group (Table). However, postprocedural diameter of the vessels differed significantly between the two groups.
|
|
|
|
|
|
|
|
The MPRI value for segments located in the area of the treated coronary artery was 1.88 ± 0.64, whereas the MPRI value in the control segments was 2.29 ± 0.47 (P < .01).
Two patients (one from the stent group, one from the PTCA only group) reported chest discomfort on exertion at the time of follow-up MR examination. These patients showed a decreased MPRI in the area of the formerly treated coronary artery of 0.82 ± 0.26, compared with 2.14 ± 0.37 in patients without recurring chest discomfort (P < .001). One patient from the PTCA only group exhibited abnormal MPRI values without evidence of clinical symptoms. In one other patient with continuing impairment of regional MPRI value immediately after PTCA, values normalized at follow-up examination.
No statistically significant difference in MPRI was found between the stent group and the PTCA only group at follow-up examination (2.20 ± 0.82 vs 2.00 ± 0.62, P = .55).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Perfusion Sequence
The motivation for using an SSFP perfusion sequence in this study was the improved image quality of perfusion images that is provided by the SSFP technique (21–24). Schreiber et al (21) found improved signal-to-noise and contrast-to-noise ratios with an SSFP perfusion sequence, which can enhance the diagnostic accuracy in patient examinations. Similarly, in studies conducted by Hunold et al (22) and Fenchel et al (24), higher signal-to-noise ratio and contrast-to-noise ratio values, as well as improved image quality, have been reported for SSFP compared with a spoiled gradient-recalled echo sequence in first-pass myocardial MR perfusion examinations. Furthermore, Chiu et al (23) successfully used an inversion-recovery SSFP perfusion sequence within a comprehensive examination protocol for the work-up of patients with acute coronary syndrome. Although Fenchel et al found increased susceptibility artifacts at the time of contrast agent inflow into the left ventricle with an SSFP perfusion sequence, artifacts could be clearly distinguished from subendocardial perfusion deficits because of their temporal and spatial characteristics (24). Therefore, SSFP perfusion techniques provide favorable signal intensity and contrast enhancement properties while exhibiting controllable artifact behavior.
Perfusion Parameters
The MPRI, derived from rest and stress examinations, has been used in several previous studies to reliably demonstrate hypoperfused myocardial regions (15,17,18,25–27). A cutoff value of 1.1–1.5 for MPRI was used in previous studies to distinguish between normal and hypoperfused regions (15,26,27). The cutoff value of 1.5 for MPRI is less than the lower normal values determined by means of different techniques in healthy control subjects (28–31), which can be explained by the reduced coronary vasodilator reserve and thus reduced myocardial perfusion reserve in coronary arteries with an increase in stenosis severity (30,32). However, Al-Saadi et al (15,25) have reported a high discriminatory power for hypoperfused myocardial regions with the use of this cutoff value. Specifically, the sensitivity, specificity, and diagnostic accuracy for the depiction of significant coronary artery stenosis were 90%, 83%, and 87%, respectively. Therefore, in the present study, the same threshold value for MPRI was chosen. MPRI values for segments supplied by stenotic or nonstenotic coronary arteries determined in our study are in good agreement with values reported previously with MR imaging (15,17,25,27), positron emission tomography (PET) (30,33,34), or Doppler ultrasonographic coronary flow reserve measurements (35,36).
Perfusion before and after Angioplasty
Residual reversible perfusion deficits after PTCA, despite normal angiographic findings, have been reported previously by using scintigraphy or other gradient-echo techniques (25,37–39). Since optimal angiographic results were seen in all patients in our study, these findings may represent microvascular circulatory impairment associated with the balloon dilation (7,40,41), general atherosclerotic processes in the epicardial vessel (6), or vasoconstriction distal to balloon injury (41).
However, in our study, MPRI normalized almost completely after stent placement. Our findings are consistent with those of a previous study by Versaci et al (38), where reversible thallium 201 (201Tl) perfusion defects were observed early after successful PTCA or stent implantation, despite residual stenosis of less than 30%. Incidence and severity of reversible 201Tl perfusion defects were increased early after PTCA compared with those after stent implantation. Similarly, Kern et al (36) found improved distal coronary blood flow and velocity after coronary angioplasty, which was, in large part, related to coronary lumen enlargement. Moreover, although some postangioplasty cross-sectional areas were equivalent to the area of a stent, the coronary vasodilatory reserve was, with rare exceptions, consistently lower after angioplasty alone. Those observations are in agreement with our results and may be caused by difficulties in appreciating the interventional success, that is, the cross-sectional area, after PTCA by means of two-dimensional projections on conventional angiograms, as reported by Gould (42) and White et al (43). This would emphasize the value of physiologically complementary methods, such as MR perfusion, for the follow-up of patients after intervention.
Eight-month follow-up MR examinations could be performed in 10 patients. This group comprised five patients who had undergone PTCA only and five patients who had undergone PTCA with stent placement. None of the patients had undergone repeated coronary angiography or SPECT examination up to the time of the MR follow-up examination. Patients not available for follow-up MR examinations were event-free up to the time of the last contact.
Versaci et al (38) reported that by using sequential 201Tl perfusion, at 6-month follow-up persistently reversible perfusion defects in the absence of restenosis at angiography were seen in a few patients treated with PTCA but in no patients treated with stent placement. Therefore, regional perfusion defects early after stent placement may have functional rather than organic causes. Three of five patients who had undergone PTCA without angiographically apparent restenosis still had regional perfusion defects (38). In contrast, in our study we observed no significant difference at follow-up examinations between patients who had undergone stent placement and patients who had undergone only PTCA. This may be due to the relatively small patient population examined at follow-up examination. Another reason may be changes in stenosis geometry for late normalization of coronary flow reserve after PTCA (9).
Serruys et al (1) reported that measurement of coronary flow reserve after PTCA, in combination with the percentage of luminal diameter stenosis, has a predictive value, albeit moderate, for the short- and long-term outcomes after PTCA and thus may be used to identify patients who will or will not benefit from reintervention. This is reflected in the two patients in our study who complained of chest discomfort on exertion and had a decreased MPRI at MR examinations. Presumably, those patients would benefit from reinterventional angioplasty procedures.
SPECT Examinations
Today, SPECT and PET still represent the reference standard for myocardial perfusion. Although several investigators have used coronary angiography as the reference for MR perfusion data (17,23,27), in the present study the standard of reference used for myocardial perfusion was 99mTc tetrofosmin SPECT, which is more frequently conducted at our institution than is PET. Limitations of SPECT perfusion and viability examinations are commonly known and are described in detail elsewhere (13,44). However, as coronary angiography enables us to determine solely the grade of epicardial vessel stenosis without assessing the functional relevance of luminal obstruction or presence of collateralization with respect to end-organ perfusion, SPECT seems to represent a more reliable reference for MR perfusion examinations (13,14,16,45). Lauerma et al (13) reported a close correlation between dipyridamole 201Tl scintigraphy and contrast-enhanced multisection MR imaging in the detection and sizing of regional myocardial perfusion defects. Furthermore, Lauerma et al suggested that multisection MR perfusion stress testing could be used to assess the effect of revascularization on regional myocardial perfusion.
Limitations
First, a relatively small number of patients were included in this study. However, despite this limitation, several statistically significant results were obtained. Studies in more patients are needed to confirm these data.
A second limitation was the determination of perfusion by means of semiquantitative perfusion parameters. After intervention, perfusion deficits could be observed in remote areas of the left ventricle. This is especially true in patients in whom a significant improvement of perfusion in formerly hypoperfused areas was observed. Supposedly, the perfusion deficits observed in areas without significant stenosis are caused by the semiquantitative nature of our data evaluation. As soon as normal blood flow is reestablished in the formerly malperfused area, other myocardial regions supplied by less stenosed coronary arteries become relatively hypoperfused.
Furthermore, serial quantification of myocardial perfusion is subject to several limitations: Myocardial perfusion varies in concordance with patient hemodynamic parameters, metabolic states, and response to pharmacologic treatment (37). Moreover, presence of reversible perfusion deficits within remote territories may be caused by early coronary artery dysfunction or microvascular disease undetectable at angiography (34). The size of perfusion beds may also be altered in relation to changes in obstructive significance of coronary lesions after angioplasty. Kaul et al (46) reported that, because of a dynamic interplay between anterograde and collateral blood flow, the size of the area perfused by a coronary artery changes during myocardial ischemia. Last, a complication of the intervention, that is, obstruction of small vessels after angioplasty, may be responsible for persisting perfusion deficits despite documented interventional success. However, in patients in this study, there was no elevation of creatine kinase values after angioplasty.
In conclusion, the MPRI, derived from rest-stress examinations, can provide information regarding the success of interventional procedures in stenosed coronary arteries. Therefore, rest-stress perfusion MR imaging may be a valuable tool in the follow-up of patients after coronary intervention.
|
FOOTNOTES |
|---|
Abbreviations: MPRI = myocardial perfusion reserve index • PTCA = percutaneous transluminal coronary angioplasty • SSFP = steady-state free precession
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, M.F., S.M., A.F., C.D.C.; study concepts, all authors; study design, M.F., S.M., A.F., U.H.; literature research, M.F., S.M., A.F., U.K., N.I.S., U.H.; clinical studies, M.F., A.F., U.K., N.I.S., U.H., S.M.; data acquisition, M.F., S.M., A.F., U.K., N.I.S., U.H.; data analysis/interpretation, M.F., S.M., A.F., U.H.; statistical analysis, M.F., S.M., A.F., U.K., N.I.S.; manuscript preparation, M.F., A.F., S.M., U.K., U.H.; manuscript definition of intellectual content, M.F., S.M., A.F., C.D.C.; manuscript editing, revision/review, and final version approval, all authors
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Gebker, C. Jahnke, I. Paetsch, B. Schnackenburg, S. Kozerke, A. Bornstedt, E. Fleck, and E. Nagel MR Myocardial Perfusion Imaging with k-Space and Time Broad-Use Linear Acquisition Speed-up Technique: Feasibility Study Radiology, December 1, 2007; 245(3): 863 - 871. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
