| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Departments of Radiology, Section of Cardiovascular Imaging (R.M.S., S.S.H., A.E.S., R.D.W.), and Thoracic and Cardiovascular Surgery (N.G.S., J.F.S., R.D.W.), the Cleveland Clinic Foundation (HB6), 9500 Euclid Ave, Cleveland, OH 44195; and Siemens Corporate Research, Princeton, NJ (T.P.O.). Received February 6, 2004; revision requested April 13; revision received November 22; accepted December 30. Address correspondence to R.D.W. (e-mail: whiter{at}ccisd1.ccf.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: This study was HIPAA compliant and was approved by the local Institutional Review Board, with waiver of informed consent. Twenty-six patients (19 men, seven women; age, 56 years ± 12 [± standard deviation]) with CIHD underwent MR imaging assessment of myocardial viability and multi–detector row CT assessment of CAD on the same day. For coregistration, a population-based LV model was fit to each data set separately; models were then registered spatially. For data analysis, correspondence between coregistered displays and the 17-segment LV model for assessment of CIHD was evaluated, accuracy of using coregistered displays to evaluate the extent of CAD and myocardial disease was assessed, and utility of coregistered displays in optimizing surgical revascularization planning was determined.
RESULTS: Coronary assignment for coregistered displays and the 17-segment LV model differed in 17% of myocardial segments. For the majority of patients, three segments (midanterolateral [62%], apical lateral [73%], and apical inferior [58%]) were discordant. Segments were supplied by the left anterior descending artery, a diagonal branch, or a ramus intermedius with diagonal distribution in all but one case. Coregistered displays were deemed concordant with selective coronary angiography and alternate myocardial imaging in all cases. Overall, surgical planning was potentially enhanced in 83% of cases because, compared with alternate imaging modalities, coregistered displays were believed to demonstrate the relationship between coronary arteries and underlying myocardial tissue more definitively and efficiently (for patients in whom surgery was performed) or more correctly and comprehensively (for a presumably better-tailored surgery).
CONCLUSION: Assessment of CIHD can be improved by using coregistered displays that directly relate the condition of LV myocardium to the anatomy of the coronary arteries in individual patients.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Acute necrosis or chronic scarring of the LV myocardium can be directly visualized by using delayed enhancement MR imaging (3,4). Furthermore, delayed enhancement MR imaging can be combined with other MR imaging techniques, such as those that enable assessment of LV wall motion and intramyocardial mechanics or myocardial perfusion, to differentiate between regions of myocardium that are normal, inducibly ischemic, chronically ischemic (eg, hibernating), or infarcted (5).
Multi–detector row CT coronary angiography is capable of depicting the coronary arteries with an in-plane spatial resolution that is approaching that of conventional angiography. Multi–detector row CT angiography also permits the visualization of major epicardial coronary arteries to their distal third, with increasingly consistent visualization of branch vessels (6). CT angiography, along with enabling detection of clinically important coronary stenoses (7), can depict mildly stenotic early atherosclerotic changes that are predisposed to instability (8,9).
The concept of using an integrated imaging approach for the simultaneous visualization of MR imaging myocardial viability maps and multi–detector row CT coronary angiography displays, which are spatially and temporally coregistered and provide a direct link between the condition of myocardial regions and the specific coronary anatomy of an individual with CAD, has been previously introduced (10). It has been hypothesized that these coregistered MR imaging and multi–detector row CT displays could improve therapeutic planning, especially in patients who are being considered for revascularization to treat chronic ischemic heart disease (CIHD).
Thus, the goals of this study were (a) to evaluate the assignment of LV myocardial segments to coronary arterial territories by using coregistered MR imaging and multi–detector row CT displays, (b) to assess the accuracy of coregistered displays in determining the presence and distribution of clinically important CAD and the regional effect of CAD on LV myocardium, and (c) to determine the potential utility of coregistered displays in optimizing surgical revascularization planning.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The study population consisted of 26 patients (19 men, seven women; age, 56 years ± 12 [± standard deviation]) with known CIHD who were referred for both MR imaging assessment and multi–detector row CT evaluation for the aforementioned reasons as part of clinical planning for possible surgical revascularization. These patients were imaged between October 1999 and December 2002; other patients with CIHD who were imaged within this period were not eligible for inclusion in this study because they were not clinically evaluated with both modalities. In all patients, MR imaging and multi–detector row CT were performed on the same day. In addition, all patients had previously undergone selective coronary angiography within 28 days ± 29 of MR imaging and multi–detector row CT, with no coronary events noted during that interval. No patients had previously undergone revascularization procedures, including bypass grafting, balloon angioplasty, or stent placement.
Data from all 26 patients were included in the analysis of goal 1, which was to evaluate the correspondence between coregistered imaging displays and a 17-segment LV model for the assignment of myocardial regions to a specific coronary artery.
Data from 19 of 26 patients were included in the analysis of goal 2, which was (a) to assess the potential utility of coregistered displays in determining the presence and distribution of clinically important CAD and the regional effect of CAD on LV myocardium and (b) to compare the utility of coregistered displays with that of selective angiography and alternate myocardial imaging. In these 19 patients, digital selective coronary angiograms, as well as data from other myocardial imaging modalities such as echocardiography (19 patients) and metabolic and perfusion PET (four patients), were available for review.
Last, data from the 15 patients who eventually underwent cardiac surgery subsequent to MR imaging and multi–detector row CT were included in the analysis of goal 3, which was to determine the potential utility of using coregistered displays to optimize the planning of surgical revascularization. Coronary artery bypass grafting was performed in 14 (93%) of these 15 patients, with an average of 3 grafts ± 1 per patient. The remaining patient (patient 7) underwent only mitral valve repair and LV reconstruction for postinfarction aneurysm (11).
MR Image Acquisition
MR imaging was performed by using a 1.5-T imager (Sonata; Siemens Medical Solutions, Erlangen, Germany). Cine images were acquired in the short-axis of the LV myocardium in contiguous sections that spanned from the mitral valve to the LV apex (balanced steady state free precession; 4.1/2 [repetition time msec/echo time msec]; flip angle, 60°; section thickness, 8–10 mm; field of view, 300–360 mm; rectangular field of view, 80%–100%; initial imaging matrix, 256 x 256). Approximately 20 minutes after intravenous injection of 0.2 mmol/kg gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ), short-axis delayed enhancement MR imaging was performed by using an inversion-recovery T1-weighted segmented gradient-echo pulse sequence (8/4/175–275 [repetition time msec/echo time msec/inversion time msec]; flip angle, 30°; and 23 lines acquired every other R-R interval, with field of view and matrix values matching those of the cine acquisition) (3,5). The optimal inversion time to null the signal from viable myocardium was determined individually for each patient. In addition, long-axis cine and delayed enhancement MR images were acquired in vertical (two chamber), LV outflow tract (three chamber), and horizontal (four chamber) orientations.
Multi–Detector Row CT Image Acquisition
Multi–detector row CT data were acquired by using either a four–detector row scanner with a 500-msec gantry rotation time (Sensation 4; Siemens Medical Solutions) (n = 17) or a 16–detector row scanner with a 420-msec rotation time (Sensation 16; Siemens Medical Solutions) (n = 9). Contrast-enhanced retrospective electrocardiographically gated spiral techniques were used in all patients. Multi–detector row CT images were reconstructed at 55% of the cardiac cycle.
Patients had a mean heart rate of 67 beats per minute (range, 47–97 beats per minute), which was measured either with (n = 4) or without (n = 22) intravenous ß-blocker administration (metoprolol, Lopressor; Novartis Pharmaceuticals, East Hanover, NJ) prior to the procedure. Contrast agent transit time was determined by using a 20-mL timing bolus (Ultravist 300; Berlex, Wayne, NJ). Intravascular enhancement of the diagnostic scan was achieved by using a single injection of 100–150 mL of the same contrast agent at 2.5–3.5 mL/sec, depending on the type of scanner used. Gated spiral CT scanning was used to image the entire heart from the base to the apex during a single breath hold. Imaging parameters for the four–detector row scanner were tube voltage, 120 kV; tube current, 300 mA; temporal resolution, 125–250 msec; four detector rows used; section thickness, 1 mm; reconstruction kernel, medium sharp body kernel (B30f); and spatial resolution, 12 line pairs per centimeter (cutoff). Imaging parameters for the 16–detector row scanner were tube voltage, 120 kV; tube current, 370 mA; temporal resolution, 105–210 msec; 12 detector rows used; section thickness, 0.75 mm; reconstruction kernel, medium sharp body kernel (B30f); spatial resolution, 12 line pairs per centimeter (cutoff).
Image Processing
Prior to image coregistration, three image postprocessing steps were required: MR image processing, multi–detector row CT image processing, and coronary artery segmentation.
For delayed enhancement MR images, the LV myocardium was segmented by manually delineating the endocardial and epicardial borders in short-axis and long-axis images with cardiovascular image processing software (Argus; Siemens Medical Solutions). This step was accomplished jointly by two experienced investigators (R.M.S. and R.D.W, with 9 and 20 years experience in cardiac MR imaging, respectively).
Multi–detector row CT image processing required two steps: reformatting and segmentation. The multi–detector row CT data set was reformatted along the cardiac axes, thereby resulting in sets of short-axis and long-axis images. Although multi–detector row CT short-axis images were parallel to delayed enhancement MR short-axis images, the multi–detector row CT short-axis images were not necessarily coplanar and were needed only for coregistration (described later). Next, the LV myocardium was segmented manually by using image processing software (Argus; Siemens Medical Solutions). This step was accomplished jointly by two investigators (R.M.S. and R.D.W., with 4 and 8 years experience in cardiac multi–detector row CT, respectively).
For the last image processing step, the coronary tree was manually segmented from the multi–detector row CT coronary angiography data by using specialized software (modified VesselView; Siemens Medical Solutions). The coronary arteries were identified by an experienced cardiovascular imager (R.D.W., 20 years experience in cardiac imaging) who delineated the course of the major epicardial coronary arteries and branch vessels by marking points on the raw transaxial images or oblique multiplanar reconstructions. These points were then saved for later inclusion with coregistered images.
During coronary artery segmentation, locations were noted for which clear truncation of the vessel lumen, which was caused by subtotal or total obstruction, was demonstrated.
Image Coregistration
Coregistration of MR imaging and multi–detector row CT data was performed in four steps. First, a three-dimensional LV average shape model (12), which consisted of endocardial and epicardial surfaces, was fit to the multi–detector row CT short-axis and long-axis image contours. The initial model rotation was controlled by specifying the right ventricle insertion points on a midventricular multi–detector row CT short-axis image. Adjustment of the gross translation and the rotation of the model, as well as of the deformation of each model surface to the data, was performed algorithmically by using a physically motivated paradigm (13). The model was capable of deforming to the contours in three ways. It could (a) rigidly translate and rotate, (b) adjust its global shape parameters (eg, wall thickness, distance from the apex to the base, and blood pool depth), or (c) act as a mesh of springs to conform to local characteristics of the LV contours after the global shape had settled (14). Although the model was able to adjust its global shape parameters, the model was constrained to a shape resembling an LV segment as these parameters were being adjusted. This model was referred to as the multi–detector row CT model.
Second, the multi–detector row CT model was copied, and the copy was fit to the delayed enhancement MR image contours by using the approach described earlier. This model was referred to as the MR imaging model.
Third, in an attempt to minimize the root mean square differences of the average distance between the two model surfaces, the surface of the multi–detector row CT model was aligned with that of the MR imaging model by using translation and rotation only.
Fourth, translation and rotation from the third step were applied to the multi–detector row CT images and segmented coronary tree to align them with the delayed enhancement MR images.
Image processing and coregistration required approximately 40 minutes per patient, and the image coregistration steps accounted for approximately 10 minutes of this time. In all patients, the adequacy of coregistration was assessed qualitatively by visual inspection of the correspondence between reformatted multi–detector row CT and delayed enhancement MR images after coregistration. Evaluation was accomplished jointly by three investigators (R.D.W., R.M.S., and A.E.S., with 20, 9, and 15 years experience in cardiac imaging, respectively).
Data Analysis
Data analysis was performed in three phases that reflected each of the study goals. Phase 1 involved single modality evaluation of regional LV myocardial viability and function by using MR imaging and a 17-segment model. Phase 1 also included direct assessment of the relationship between regional viability, function, and coronary anatomy on the basis of coregistered MR imaging and multi–detector row CT displays. Phase 2 involved assessment of the accuracy of coregistered displays compared with that of selective coronary angiography and alternate modality myocardial imaging (eg, echocardiography). Phase 3 involved assessment of the utility of coregistration for enhanced surgical revascularization planning on the basis of postsurgical retrospective review of preoperative, intraoperative, and postoperative data.
To evaluate the condition of the regional myocardium by using MR imaging (phase 1), an experienced cardiovascular imager (R.D.W.) who was blinded to the condition of the coronary arteries reviewed the matched delayed enhancement and cine images by using a 17-segment LV model. On delayed enhancement MR images, LV segments were graded according to the extent of myocardial scarring—that is, no scar, <50% scar, or 
50% scar. Segmental scar grading and the 50% scar threshold for likelihood of improved function following successful surgical revascularization have been previously described (15). In cine loops that were matched to the delayed enhancement MR images, segmental wall motion was graded as normal, hypokinetic, or akinetic or dyskinetic.
For phase 1, the condition of the LV myocardium was assessed by using coregistered data and was evaluated relative to data obtained with the 17-segment model. To do this, the reviewer viewed the segmented coronary arteries that were superimposed on the MR images and directly assigned each LV myocardial segment to a coronary artery system.
For each of the 19 patients included in phase 2, the accuracy of coregistered MR imaging and multi–detector row CT interpretations was separately assessed by two experienced cardiac surgeons (N.G.S. and J.F.S., with 15 and 14 years experience, respectively) who were blinded to the previous image processing steps but had access to the final coregistration results and other imaging data, including conventional coronary angiograms. Each surgeon used all available clinical data, including the reported surgical results and immediate postoperative findings, as well as imaging data, including both digital angiographic data and clinical reports from cardiologists. The clinical reports of routine, subjectively assessed examinations were used by surgeons to develop their impressions of the pre- and postrevascularization conditions of each patient. A binary accuracy score was then assigned to each patient. A score of 1 indicated that there was agreement with the angiographically defined coronary artery distribution and that clinically important coronary stenoses and myocardial abnormalities were accurately defined on coregistered displays. A score of 0 indicated that coregistered results were not confirmed with angiography or other myocardial imaging modalities.
Last, for each of the 15 patients included in phase 3, the potential effect of coregistered data on surgical revascularization planning (phase 3) was separately judged by the same two cardiac surgeons (N.G.S., J.F.S.) by using an "impact score." A score of 0 meant that there was no additional information provided by the coregistered data. A score of 1 indicated that the coregistered data provided additional information for enhanced surgical planning but that the resulting surgery would not have changed. A score of 2 meant that the coregistered data provided a large amount of additional information that could be used to enhance surgical planning, as well as to potentially optimize surgical revascularization.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Evidence of incomplete or complete coronary artery occlusion was noted in 12 (46%) of 26 patients; three of these patients exhibited incomplete and/or complete occlusions in multiple vessels.
MR imaging.—All myocardial segments (442 total segments) were evaluated in each of the 26 patients whose images were coregistered. For the evaluation of myocardial scarring, 139 (31%) of 442 segments had no scar, 181 (41%) of 442 segments had <50% scar, and 122 (28%) of 442 segments had 50% scar. For the assessment of LV function, 109 (25%) of 442 segments were normal, 195 (44%) of 442 segments were hypokinetic, and 138 (31%) of 442 segments were akinetic and dyskinetic.
Correspondence between LV Myocardium and 17-Segment Model
Image coregistration was deemed adequate in all cases; examples are shown in Figure 1. After the relationship between the myocardial and coronary anatomy was directly assessed by using coregistered MR imaging and multi–detector row CT data, 76 (17%) of 442 segments were found to differ from the coronary artery distribution that was empirically assigned with the 17-segment model (Fig 2). Discordance between at least one segment and the supplying artery was observed in 22 (85%) of 26 patients, with an average of 2.9 ± 1.9 instances of discordance per patient (range, 0–8) (Table 1). The average frequency of discordance is shown in Figure 2 according to segment. According to coronary artery territory, 0 (0%) of 182 LAD segments, 45 (35%) of 130 LCX segments, and 31 (24%) of 130 right coronary artery segments were determined to be discordant. According to level, 15 (10%) of 156 basal segments, 27 (17%) of 156 midventricular segments, and 34 (26%) of 130 apical segments were discordant.
|
|
|
|
|
|
Utility of Coregistered Displays in Optimizing Surgical Revascularization Planning
Impact scores assigned by both surgeons are displayed in Table 2. Coregistered data that provided additional information deemed important for surgical planning corresponded to impact scores of 1 (improved planning) or 2 (potentially optimized revascularization) in 11 (73%) of 15 patients according to surgeon A and in 14 (93%) of 15 patients according to surgeon B.
|
In the six patients for whom revascularization could have been potentially enhanced had coregistered MR imaging and multi–detector row CT displays been made available during surgical planning (ie, in patients with an impact score of 2), the benefits were derived from understanding the contribution of diagonal branches to the lateral LV segments that were otherwise misassigned to the LCX with the 17-segment model. An example of this is shown in Figure 4. In another case, the additional importance of the distal branches of the dominant LCX to the inferior LV segment was deemed important to surgical planning.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
For many of the patients in this study in whom coregistration was deemed potentially beneficial to surgical planning (surgeon A, seven [64%] of 11 patients; surgeon B, 12 [86%] of 14 patients), the perceived benefits of coregistration were attributed to a discovered discordance between the supplying coronary artery that was empirically assigned to a myocardial segment by using the 17-segment model and the artery that was revealed to be responsible by using the coregistered MR imaging and multi–detector row CT display. The accuracy of the coregistration display was subsequently confirmed by using coronary angiography and alternate myocardial imaging modalities. Although the coregistered display, when compared with the 17-segment approach, consistently enhanced diagnostic confidence in these patients (and was therefore thought to enhance surgical planning), we do not believe that surgery would have been optimal as a result of these findings.
In other patients (surgeon A, four [36%] of 11 patients; surgeon B, two [14%] of 14 patients), however, we determined that image coregistration could have had a major positive effect on surgical planning and potentially altered the specific placement and number of bypass grafts. In patients 1, 4, 5, and 10 for surgeon A and patients 3 and 13 for surgeon B, image coregistration was used to identify large diagonal branches of the LAD that supplied territories otherwise empirically assigned to the LCX. Surgeon A believed that the addition of bypass grafts to the pertinent diagonal branches of the LAD would have been advantageous in three patients; this surgeon also believed that, in one patient, knowledge of the importance of a prominent diagonal branch could have made the vein graft to the LCX unnecessary. Surgeon B believed that two patients should have undergone bypass grafting of the diagonal branches, both of which were identified as serving lateral segments on coregistered displays and one of which was identified as serving the distal dominant LCX system owing to its newly recognized importance to the inferior LV wall.
In a previous study (18), the importance of evaluating the presence and severity of CAD prior to surgery was demonstrated by evaluating the determinants of outcome after surgical revascularization for ischemic cardiomyopathy. Results of that study demonstrated that the two most important predictors of event-free survival were the condition of the coronary arteries before surgery (a subjective measure that was based on the presence of stenosed and occluded arteries in the whole arterial tree, as viewed on coronary angiograms) and the presence of viable myocardium in target areas (18). Other statistically significant predictors, including elective surgery, complete revascularization, and the number of bypass grafts, were related to the surgical procedure itself. Hence, the most important determinants of outcome could be evaluated before surgery by using noninvasive imaging and could therefore be possibly enhanced by using the coregistration approach as described in the current investigation.
This study highlights the potential effect of image coregistration. Currently, the condition of the LV myocardium and coronary arteries must be evaluated separately, which requires physicians to mentally integrate data from disparate sources in their evaluation of disease presence and severity (10). Through the use of coregistered displays, however, the anatomic, histologic, and functional data can be visualized simultaneously on a single software platform. Furthermore, coregistration is not limited to MR imaging and multi–detector row CT; other modalities, such as PET for the evaluation of metabolism and/or perfusion and selective x-ray angiography for the evaluation of coronary arteries, could also be incorporated into the existing model.
Although variations in coronary anatomy assignment have been noted previously with the 17-segment model (2), in the current study results of only four of 26 patients were found to agree with the model in all segments; on average, patient results were found to differ in approximately three segments. LCX segments were discordant most often (45 [35%] of 130 segments), with the majority of these segments actually located within the LAD distribution. By level, apical segments were found to differ most often (34 [26%] of 130 segments), with segment 15 (apical inferior segment) differing in 15 (58%) of 26 patients and segment 16 (apical lateral segment) differing in 19 (73%) of 26 patients. Previously, the greatest variability was associated with segment 17 (apical cap) (2), although this segment was deemed concordant in all patients in the current study.
In one study (19), researchers assessed the accuracy of a different 17-segment LV model by using monoplane coronary angiography. This model differed from the model used in the current study at the apex only, with the apical third divided into three segments (instead of four) and the apical cap divided into two segments (instead of one). Similar to the current study, these investigators (19) found discordance to be most common in the lateral wall of the apex, which was supplied by the LAD in 93% of patients. In addition, they found that 80% of patients corresponded to their model in at least 14 segments; in the current study, only 65% of patients had at least 14 concordant segments.
There are few readily identifiable landmarks in the LV myocardium; however, the overall shape of the LV segment in an individual is well defined. Therefore, we employed a model-based approach as opposed to a point-based method for the task of registration. One limitation of the current study was that no objective measure of the adequacy of coregistration was employed. In spite of this, the rotation of multi–detector row CT images relative to MR images was not observed in any patients, and the apex showed good spatial correspondence between modalities. Therefore, any existing registration errors were probably restricted to the radial direction, which should not affect the assignment of coronary territories.
Another possible source of registration error occurs because of temporal differences in the acquisition and/or reconstruction of multi–detector row CT and MR images. Multi–detector row CT images were reconstructed at 55% of the cardiac cycle, as is done clinically at our institution by using an acquisition window that is longer than that used for MR imaging. MR images, however, were inside the multi–detector row CT reconstruction window in all patients. Furthermore, we believe that errors resulting from the reconstruction window would be restricted to the radial direction, which should not affect the assignment of coronary territories.
Additional limitations include the use of only one radiologist reviewer for goals 1 and 2 of this study, which could have biased the results. Also, phase 3 was performed retrospectively, although we believe this study design was necessary as a first step toward showing the clinical utility of coregistered image display in the planning of surgical revascularization.
In conclusion, implementation of techniques such as the image coregistration approach that is detailed in this study may potentially enhance clinical practice by reducing the reliance on idealized models to describe the correspondence between coronary anatomy and the underlying LV myocardium. Coregistered multi–detector row CT and MR imaging displays enhance the assessment of the relationship between the myocardial condition and the condition of the coronary arteries in the individual patient who has CAD and CIHD and can therefore be used to optimize surgical revascularization planning.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: CAD = coronary artery disease • CIHD = chronic ischemic heart disease • LAD = left anterior descending artery • LCX = left circumflex artery • LV = left ventricular
Author contributions: Guarantors of integrity of entire study, R.M.S., R.D.W.; study concepts, R.M.S., T.P.O., S.S.H., R.D.W.; study design, R.M.S., T.P.O., N.G.S., R.D.W.; literature research, R.M.S.; clinical studies, N.G.S., J.F.S., A.E.S., R.D.W.; data acquisition, N.G.S., J.F.S., A.E.S., R.D.W.; data analysis/interpretation, R.M.S., N.G.S., J.F.S., R.D.W.; statistical analysis, R.M.S.; manuscript preparation, R.M.S., T.P.O., S.S.H., R.D.W; manuscript definition of intellectual content, R.M.S., T.P.O., N.G.S., R.D.W.; manuscript editing, R.M.S., R.D.W.; manuscript revision/review, R.M.S., T.P.O., N.G.S., J.F.S., S.S.H., A.E.S.; manuscript final version approval, all authors
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. G. Danias Concordance between actual and expected coronary artery distribution. J. Am. Coll. Cardiol. Img., September 1, 2008; 1(5): 688 - 688. [Full Text] [PDF]
|
|
|
|
|
|
J. T. Ortiz-Perez and E. Wu Reply J. Am. Coll. Cardiol. Img., September 1, 2008; 1(5): 688 - 689. [Full Text] [PDF]
|
|
|
|
|
|
J. T. Ortiz-Perez, J. Rodriguez, S. N. Meyers, D. C. Lee, C. Davidson, and E. Wu Correspondence between the 17-segment model and coronary arterial anatomy using contrast-enhanced cardiac magnetic resonance imaging. J. Am. Coll. Cardiol. Img., May 1, 2008; 1(3): 282 - 293. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. T. Wilson, P. Gopalakrishnan, and T. Tak Noninvasive Cardiac Imaging with Computed Tomography Clin. Med. Res., October 1, 2007; 5(3): 165 - 171. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
). (c) Coregistered myocardial MR images and coronary CT angiograms with occluded segment of LAD (red) are superimposed on postinfarction aneurysm. Note that MR image planes coincide with those of b. Apical lateral extension of prominent second diagonal branch of LAD into regions assigned empirically to LCX with 17-segment model was considered by cardiac surgeon to be substantially beneficial, with a potentially substantial effect on surgical revascularization planning. This beneficial effect was caused by new appreciation for contribution of proximal LAD system to viable lateral regions of LV, which warranted specific revascularization. D1 = first diagonal, RCA = right coronary artery.
