| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||
ABSTRACT |
Wednesday, May 3, 1:30 PM-3:30 PM
Abstracts 174-185
Moderators: Joel E. Fishman, MD and Caroline Chiles, MD
1:30 PM
174. Statistical Reconstruction of Targeted Regions with Application to CT and PET Cardiac Imaging
Alessio Adam. M.*; Kinahan P.E.; Radiology, University of Washington, Seattle, WA.
Address correspondence to Adam.M. Alessio (aalessio{at}u.washington.edu)
Objective: Accurate cardiac imaging requires gating the cardiac cycle into individual time frames. For nuclear imaging this results in multiple low-count scans of the heart and for CT imaging this usually requires added radiation dose to the patient, or multiple low-dose scans. These low-count scans result in multiple high-noise images. Statistical reconstruction is effective at reducing noise levels in images. We propose an improved statistical reconstruction method applicable to either PET or CT imaging that is constrained by the summed immobile portions of the images and applied to only the cardiac region.
Materials and Methods: The proposed reconstruction method first performs an analytic reconstruction of the portion of the image that is stationary (region outside mediastinum) using measurements from all time frames. Then, it executes a fast statistical reconstruction of only the kinetic portion of the image for each time frame constrained by the higher-count stationary region. These steps effectively form a low-noise image of the cardiac cycle at each phase. While this method is applicable to either CT or PET imaging, we initially test its efficacy with PET simulations of cardiac motion using the NCAT digital phantom. We gated the cardiac cycle into 8 time frames and generated 8 sets of PET projection data with realistic noise levels. We positioned a circular region of interest (ROI) around the kinetic cardiac region and use all data external to this ROI for the higher-count stationary reconstruction. The data was reconstructed with the proposed method and conventional algorithms for comparison.
Results: Our complete reconstruction targeted on the cardiac ROI
required 1/3 the computation time as standard statistical reconstruction.
Along with speed improvements, the targeted ROI method reduced the mean
absolute error in the cardiac region by 
5% and the noise by 5%
averaged over all 8 time frames. Algorithm refinements of minimizing the
cardiac ROI and optimized image regularization will further reduce both the
reconstruction time and error.
Conclusion: Constraining dynamic reconstruction with the higher count, stationary portions of the scan resulted in reduced error and noise levels. Furthermore, reconstructing only a targeted ROI offers a faster algorithm for generating low-noise images of the heart for CT and PET imaging.
175. Computation of Left Ventricular Ejection Fraction with Gated Cardiac CT Angiography
Zhang S.*; Halpern E.J.; Wansaicheong G.; Levin D.C.; Radiology, Thomas Jefferson University, Philadelphia, PA.
Address correspondence to S. Zhang (shaoxiong.zhang{at}jefferson.edu)
Objective: To compare estimates of left ventricular ejection fraction (EF) obtained from gated cardiac CT angiography by subjective estimation and by two different quantitative algorithms.
Materials and Methods: Gated cardiac CT was obtained in 45 subjects with the Brilliance Pro 16-slice scanner (Philips Medical Systems). Images were reconstructed at 3.0mm slice thickness at 10% increments throughout the cardiac cycle. A subjective estimate of the EF was based upon a cine loop display including a two chamber view, a four chamber view and short axis views. The "gold standard" EF was computed by Simpson's algorithm with 14 short axis tracings of the left ventricular cavity. An additional estimate was obtained with a modified Simpson's algorithm using a biplane technique that combines the horizontal and vertical long axes. Quantitative analysis was conducted by two observers. Inter- and intra-observer variability was assessed.
Results: Estimated EF with the short axis method ranged from 34%
to 77% (mean = 61.7%). Subjectively estimated EF ranged from 45% to 80%, and
was 1.9% higher on average compared to the short axis method (p =
0.02). EF based upon the biplane techniques was 3-5% higher on average than EF
obtained by the short axis method (p < 0.01). Pearson correlation
coefficients demonstrated superior correlation with shortaxis EF for the
biplane technique (0.88) as compared to the subjective EF (0.80). The
intra-observer correlation for subjective estimate, biplane and short axes EF
were 0.73, 0.86 and 0.82 respectively. Inter-observer correlation was also
higher in biplane computation (0.71) than in short axis EF measurement (0.65).
Definition of the left ventricle in the horizontal and vertical long axes
required substantially less time (< 1 minute/patient) as compared to manual
adjustment of the automated short axis tracings ( 5 minutes/patient).
Conclusion: Gated cardiac CT angiography provides a clear definition of the left ventricular cavity during systole and diastole. The quantitative biplane computation requires only minimal additional time for tracing the left ventricular cavity and provides improved correlation with the "gold standard" short axis EF as compared to subjective estimates. It is a reproducible quantitative measure of left ventricle EF.
176. Evaluation of New Post-processing Software for Left Ventricular Functional Analysis in MDCT Coronary Angiograms Cann A.; Sirineni G.*; Pottala K.; Tigges S.; Syed M.; Kalra M.K.; Radiology, Emory University Hospital, Atlanta, GA.
Objective: Left ventricular (LV) functional analysis with MDCT is cumbersome at present, requiring substantial manual interaction and a high workflow time. These limitations restrict the widespread use of deriving LV functional information from MDCT datasets of coronary angiograms. We evaluated recently developed, dedicated cardiac analysis software, Circulation (Siemens Medical Systems), to assess the overall ease of operation, workflow, user satisfaction and accuracy in LV functional analysis.
Materials and Methods: MDCT coronary angiogram was performed on 15 patients (mean age 55.5 yrs, age range 18yrs to 73 yrs, male: female 2:1) with a 64 slice MDCT scanner (Sensation 64, Siemens Medical Solution). Images were reconstructed with B25f kernel, at 10 phases of R-R interval at 10% increments and with slice thickness of 3 mm and inter-slice distance of 3 mm. Two readers evaluated LV function with Circulation (Siemens Medical Solution). Objective evaluation of the Circulation software for calculation of LV ejection fraction was performed. Subjective evaluation of workflow, ease of operation and reliability of automation was performed using a 5 point scale (5 = excellent, 4 = good, 3 = average, 2 = poor, 1 = unacceptable). The LV ejection fractions with MDCT were compared to the findings on echocardiogram. Data were analyzed using statistical correlation tests.
Results: The average time for LV Functional analysis with MDCT was 4.72 ± 1.2 minutes. Automatic cardiac planes display for LV functional analysis with Circulation software was adequate in most cases, although some manual adjustment was required. Automatic segmentation of LV was adequate in 50% of cases but required manual adjustment by "expand and contract" function in remaining cases. The LV ejection fraction from MDCT data showed excellent correlation with echocardiography (r2 = 0.84, p < 0.017).
Conclusion: The Circulation software allows satisfactory and reliable analysis of LV function with acceptable interpretation time. The calculated ejection fraction with the new software shows excellent correlation with echocardiography.
177. ECG-gated 64-row Multidetector Cardiac CT Angiography for the Detection of Patent Foramen Ovale
Williamson E.E.1*; Araoz P.A.1; Alvarez L.2; Edmister W.B.1; Borgeson D.K.2; Khandheria B.K.2; Breen J.F.1; 1. Diagnostic Radiology, Mayo Clinic, Rochester, MN; 2. Cardiology, Mayo Clinic, Rochester, MN.
Address correspondence to E.E. Williamson (ewilliamson{at}mayo.edu)
Objective: 1) Describe the CT features of patent foramen ovale (PFO), and 2) Determine the sensitivity and specificity of CT for the diagnosis of PFO, using trans-esophageal echocardiography (TEE) as a reference standard.
Materials and Methods: A retrospective chart review was performed on the first 214 consecutive patients referred for a clinically-indicated, 64-row multidetector coronary CTA and 20 patients were found who previously had a transesophageal echocardiogram (TEE). CT criteria were established for the diagnosis of PFO and blinded consensus review of each MDCTA was performed by two experienced observers. The results of this review were compared to TEE, which served as a reference standard.
Results: CT criteria for the diagnosis of PFO were 1) presence of a distinct "flap" in the left atrium in the expected location of the septum primum, 2) presence of a continuous column of contrast connecting this flap to the right atrium, and 3) presence of a "jet" of contrast from the column into the right atrium. Of the 20 patients that had both a TEE and a cardiac CTA, 6 (30%) were found to have a PFO by TEE. Using the presence of a left atrial flap as the primary diagnostic criterion, all 6 cases of PFO were detected using CT (sensitivity = 100%). Of the 14 patients with no PFO seen on TEE, 12 of these were correctly identified using CT (specificity = 86%).
Conclusion: PFO is a common congenital cardiac abnormality which is a clinically-important cause of cryptogenic stroke, particularly in young patients. The use of cardiac CT angiography for the detection of PFO has not been previously described. ECG-gated cardiac CT angiography performed with a 64-detector CT scanner can be used to reliably detect patent foramen ovale. This technique could provide a non-invasive means of screening patients at risk for PFO and assist with therapy planning.
178. Is Detection of Abnormalities Other than Fat More Important in Diagnosing Arrhythmogenic Right Ventricular Dysplasia on MRI? A Segment-based Analysis of the Right Ventricle
Gulati G.S.*; Juneja R.; Ahuja M.C.; Sharma S.; Cardiac Radiology, All India Institute of Medical Sciences, New Delhi, Delhi, India.
Address correspondence to G.S. Gulati (gulatigurpreet{at}rediffmail.com)
Objective: To test the hypothesis that detection of abnormalities other than fatty infiltration may be more reliable in diagnosing arrhythmogenic right ventricular dysplasia (ARVD).
Materials and Methods: This was a retrospective cross-sectional case-control analysis wherein a 1.5 T MRI was performed in 23 ARVD cases (based upon Task Force Criteria and not including MRI findings) and 10 controls (non-arrhythmia indications). Imaging consisted of fast spin echo T1 weighted and steady state free precession (truFISP) cine imaging, both in true axial and cardiac short axis planes. The right ventricle (RV) was divided into 10 segments (from right to left on axial images: segments 1,2 and 3 are the lower and 4,5 and 6 are the basal, mid, and apical inflow segments, 7 and 8 through the outflow tract, and 9 and 10 through the inferior portion of RV). In each segment, presence of fat, anterior myocardial thickness, aneurysm and scallop formation, systolic thickening, wall motion abnormality (WMA) and average end-diastolic diameter were evaluated. Student's t-test (continuous variables) and Chi-square test (categorical variables) were employed to compare the 2 groups (p value significant at < 0.05).
Results: Three cases were excluded (claustrophobia or suboptimal image quality). Among cases, there were 16 males and four females (mean age 36.7, range 19-58 yrs). Segment 7 was most frequently abnormal (presence of fat, thin myocardium, aneurysm/scallop formation, absent systolic thickening and WMA). WMA was the most frequent finding (111/200 segments). There was no significant difference for the presence of fat in any segment between the 2 groups (overall, fat present in 12/200 segments in cases, 4/100 segments in controls). Although 11 segments among cases showed aneurysm/scallop formation compared to none among controls, this difference did not approach statistical significance. Among cases compared to controls, the end-diastolic diameter was significantly greater for all segments, and except segments 3, 8 and 9, the anterior RV myocardium was significantly thinner. Except segments 3, 9, and 10, the p value was < 0.05 for the presence of WMA and absence of systolic thickening among cases compared to controls.
Conclusion: Fatty infiltration on MRI is infrequently observed in patients with ARVD, and may be seen in normal subjects. Its role along with presence of other structural myocardial abnormalities may have been overstressed. Presence of functional abnormalities may be more important for diagnosing ARVD.
179. Wayward Leads: Prevalence and Characterization of Pacemaker and ICD Lead Perforation on CT
Hirschl D.*; Jain V.R.; Haramati L.B.; Spindola-Franco H.; Gross J.A.; Radiology and Cardiology, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, NY.
Address correspondence to D. Hirschl (dhirschl{at}aecom.yu.edu)
Objective: Pacemakers and ICDs are widely used for the management of cardiac tachy and brady arrhythmias. Acute complication rate from device implantation is 3-7%. Other complications occur, but may be unrecognized. Lead perforation is often asymptomatic, and its incidence and significance is uncertain. The aim of the current study is to describe the incidence of pacer and ICD lead perforation on CT and to correlate the CT findings with objective device data.
Materials and Methods: We retrospectively identified 82 consecutive patients with permanent pacemakers or ICDs imaged on a 16-slice CT scanner by reviewing chest CT reports. There were 44 men and 38 women, mean age 72 (23-92). Mean interval between device placement and CT was 53 months. The images were reviewed by 2 of 3 cardiothoracic radiologists in consensus. If there was disagreement, a majority (2/3) reading was obtained. Each CT was reviewed for device type, fixation type (active vs. passive), tip position and presence of pericardial effusion. Results were correlated with lead impedance and pacing threshold, when available [80% (66/82)]. Lead fixation type was characterized by device model in 80% (66/82) and by radiologic appearance in 20% (16/82). A cardiac electrophysiologist interpreted device data.
Results: All 82 patients had right ventricular leads (46 passive, 36 active) and 49 had right atrial leads (11 passive, 38 active). 17% (14/82) of patients had a lead perforation. Perforation rates were 18% (9/49) for atrial and 6% (5/82) for ventricular leads (p < 0.05, chi square). 67% (6/9) of atrial and 60% (3/5) ventricular perforated leads had active fixation (p = NS). 21% (3/14) with perforated and 19% (13/68) with non-perforated leads had a pericardial effusion (p = NS). All effusions, but one (non-perforated lead), were fluid attenuation. Lead impedance and pacing thresholds did not differ significantly between perforated and non-perforated atrial and ventricular leads.
Conclusion: Lead perforation was fairly common on CT. Atrial leads perforated more frequently than ventricular leads. Lead impedance and pacing threshold did not differ significantly between perforated and non-perforated leads.
180. Dual Flow Contrast Injection for Coronary CTA Improves Visualization of the Right Heart
Jensen C.T.*; Ravenel J.G.; Nguyen S.A.; Costello P.; Schoepf U.J.; Radiology, Medical University of South Carolina, Charleston, SC.
Address correspondence to C.T. Jensen (Jensenct{at}musc.edu)
Objective: A saline chaser is ordinarily used at coronary CTA for contrast savings, homogenous attenuation and reduced streak artifacts. In many cases, however, the void of contrast precludes right heart analysis. We aimed at improving right heart visualization by automated injection of a contrast/saline mixture during the second phase of injection.
Materials and Methods: 15 patients were scanned with a monophasic, iodine-only protocol using a single-syringe injector; 15 patients with a biphasic protocol (iodine bolus followed by a saline chaser) using a dual-syringe injector; and 15 patients with a triphasic protocol with Dual Flow (Medrad) technology with simultaneous injection from two syringes to achieve a desired mixing ratio of contrast and saline. The iodine bolus was followed by a 50:50 saline/contrast mixture and a saline chaser. One radiologist rated the visualization of right and left heart structures (heart valves, moderator band, etc.) and the degree of artifacts. One observer performed attenuation measurements of the left and right heart and of the coronary arteries.
Results: Contrast attenuation (mean±SE) in the right heart was significantly [p = 0.009, One-Way ANOVA] lower in the biphasic group (173.2±38.5) than in the monophasic (317.1±43.3) and Dual-Flow groups (323.3±27.6). For the coronary arteries, there were no significant differences between the branches (proximal, middle, and distal) [p > 0.05, Duncan's multiple comparison]. Right heart structures were rated significantly [p < 0.05] better and artifacts occurred less frequently in the Dual Flow group, while left heart structures showed no difference.
Conclusion: Dual flow injection provides sufficient enhancement for assessment of the right heart while streak artifacts from dense contrast material can generally be avoided. Thus, this approach appears recommendable for coronary CTA, as right heart pathology (thrombo-emboli, tumors, etc.) can be reliably detected.
181. Evaluation of Anomalous Coronary Arteries with 16- and 64-Slice Multidetector CT
Patel S.1*; Bogot N.2; Kazerooni E.1; Sundaram B.1; Moscucci M.1; Chetcuti S.1; Durst R.2; 1. University of Michigan, Ann Arbor, MI; 2. Hadassah Hebrew University Medical Center, Jerusalem, Israel.
Address correspondence to S. Patel (smitap{at}umich.edu)
Objective: Anomalous coronary arteries may cause myocardial ischemia or sudden death. We reviewed the accuracy of multidetector CT (MDCT) in comparison to coronary angiography for the evaluation of anomalous coronary artery origin, and malignant course between the aorta and pulmonary outflow tract.
Materials and Methods: 300 patients undergoing ECG-gated coronary CT angiograms using 16 or 64-slice MDCT were reviewed for the presence of anomalous coronary arteries. Ostial origin and anatomic course with respect to the aorta and pulmonary artery outflow tract were analyzed. Coronary angiography served as the reference standard.
Results: 20 of 300 (6.6%) patients had anomalous coronary arteries (M = 10, F = 10, mean age 56.8 years). Ten patients had a single coronary artery origin from the aorta, 9 patients had two coronary arteries arising from the aorta, and 1 patient had 3 arteries arising from the aorta. Anomalies involved the origin of the left main coronary artery in 8 patients (40%), right coronary artery in 3 patients (15%), left circumflex coronary artery in 7 patients (35%) and left anterior descending in 2 patients (10%). Nine patients (45%) had an interarterial "malignant" course; 7 involved an anomalous left main coronary artery (6 arising from right coronary artery; 1 arising from right coronary sinus) and 2 anomalous right coronary arteries arose from the left main coronary artery. Anomalous coronary artery ostia and course were confirmed at angiography (20/20) and surgery (8/10).
Conclusion: 16 and 64-row MDCT accurately detects the origin and course of the anomalous coronary arteries.
182. Systematic Review of 16-slice Coronary CTA Performance: Eliminating One Source of Bias
Carlos R.C.*; Patel S.; Kazerooni E.A.; Radiology, University of Michigan, Ann Arbor, MI.
Address correspondence to R.C. Carlos (rcarlos{at}umich.edu)
Objective: Recent studies highlight the relatively high sensitivity and specificity of 16-slice coronary CTA (CCTA) compared to conventional coronary angiography (CCA). However, these reports exclude data on patients in whom the images are deemed inadequate for interpretation due to technical factors. This method of analysis overestimates the performance of CCTA. We sought to determine the lower bounds of CCTA performance accounting for uninterpretable results bias using systematic review.
Materials and Methods: Five peer-reviewed articles identified on Medline using standard search criteria described five separate prospective observational studies, where all patients underwent both CCTA and CCA. All studies categorized significant disease as = 50% segmental stenosis. A single reviewer extracted the total number of patients enrolled and total number of patients included in the analysis in each study. Patient level data were pooled and descriptive statistics (sensitivity and specificity of CCTA) were calculated accounting for patients who were excluded for technical factors. By initially treating excluded patients as false positives, the lower bound of specificity was calculated. Then treating excluded patients as false negatives, the lower bound of sensitivity was calculated. Similar analyses were conducted at the level of the coronary segments, accounting for segments that were excluded for technical factors.
Results: Patient level data could be extracted from 4 of the 5 studies. 227 patients were recruited, 225 were included in the reported analyses. Pooled sensitivity and specificity in the reported patients were both 85% for detection of any significant stenosis. After inclusion of all 227, the lower bounds of sensitivity and specificity dropped modestly to 83%. Coronary segment level data could be extracted from all 5 studies. A total of 2608 possible segments could be included, 2469 were ultimately analyzed and reported in the literature. Pooled sensitivity and specificity of CCTA in detection of significant stenosis in the reported coronary segments were 82% and 96%. After including all possible segments, the lower bounds of sensitivity and specificity dropped to 60% and 90%, respectively.
Conclusion: The ability of CCTA to detect any significant stenosis at the level of the patient appears robust, with minimal effect of uninterpretable results bias. However the sensitivity of significant stenosis detection at the level of the coronary segments is more susceptible to this bias.
183. Effect of Contrast Concentration, Tube Potential and Reconstruction Kernels on MDCT Evaluation of Coronary Stents: An In Vitro Study
Sirineni G.K.*; Pottala K.M.; Kalra M.K.; Cann A.D.; Syed M.; Waldrop S.; Tigges S.; Department of Radiology, Emory University School of Medicine, Atlanta, GA.
Address correspondence to G.K. Sirineni (sgkreddy{at}yahoo.com)
Objective: Visualization of coronary in-stent lumen on coronary 64 slice MDCT angiography (CCTA) is inadequate due to blooming and beam hardening artifacts from metallic stents. We evaluated the relationship of different kVp, reconstruction kernel and contrast concentrations on stent luminal diameter.
Materials and Methods: Two coronary stents with lumen diameter of 2.75 mm and 3.0 mm were deployed in silicone tubes. Tubes were filled with diluted iodinated contrast (1:20 dilution of Iohexol 350 mg% to give an attenuation value of 550 HU). These tubes were scanned at 80,100,120 and 140 kVp (n = 4). Each scan acquisition was reconstructed using B10f, B25f, B31f, B36f, B41f, B46f, B60f, and B80f kernels (higher numbers indicating a higher spatial frequency) (n = 8). The above procedure was repeated using 1:35 contrast dilution to obtain contrast attenuation value of 350 HU. Luminal diameter was measured at mid stent level for each stent, in datasets acquired at different kVp, contrast concentrations, and reconstruction kernels (n = 64 datasets*2stents = 128). In addition, luminal attenuation values (HU) were noted at the mid stent level and at a distance of 1cm from the stent within the tube lumen. One way ANOVA test and regression analysis were used for statistical analysis.
Results: The change in kernel significantly affected the difference in luminal HU values at mid stent level and away from the stent (p < 0.001), with B46f showing the least difference in HU values. The lower contrast concentration (350 HU) showed substantially less artifactual stent stenosis compared to high contrast concentration (550 HU) (p < 0.001).
Higher kernel (B46f, B60f, and B80f) substantially improved the measurement of stent luminal diameter (p < 0.001). However, kVp did not have significant effect on the visualization of stent luminal diameter (p > 0.277). There was excellent interobserver agreement for stent luminal diameters and attenuation value measurements (r2 = 0.971, p < 0.001).
Conclusion: Reconstruction kernels and contrast concentration (HU)
have significant effect on visualization of in-stent luminal diameter and
artifactual stenosis. In clinical practice, B46f kernel and lower contrast
enhancement value ( 350 HU) may be useful for evaluating the stent lumen.
Interestingly, kVp variation has no effect on artifactual stenosis of stent
lumen.
184. Can Assessment of Perfusion Defects on Multidetector CT in Post Acute Myocardial Infarction (AMI) Patients help Predict Myocardial Functional Recovery?
Ghersin E.1*; Lessick J.2; Litmanovich D.1; Dragu R.; Mutlak D.; Rispler S.2; Kopeliovitz M.2; Hammerman H.2; Agmon Y.2; Beyar R.2; Engel A.1; 1. Diagnostic Imaging, Rambam Medical Center, Haifa, Israel; 2. Cardiology, Rambam Medical Center, Haifa, Israel.
Address correspondence to E. Ghersin (e_ghersin{at}rambam.health.gov.il)
Objective: To evaluate the relationship between presence and characteristics of myocardial perfusion defects (MPD) on multidetector CT (MDCT) and myocardial functional recovery (MFR) as assessed by echocardiography, in patients following acute myocardial infarction (AMI).
Materials and Methods: Gated CT angiography was performed on 34 patients, admitted with AMI, within a mean interval of 4.8 days from admission, using a Brilliance 16 scanner (Philips Medical Systems, Cleveland). Segmental dysfunction was evaluated by echocardiography during the initial hospitalization and following 2-4 months, using a 16-segment approach. MDCT data sets were assessed for the presence, position and size of MPD, by a radiologist, blinded to the clinical diagnosis and to results of echocardiography. A MPD was defined using the following criteria: enhancement > 20 HU below normal myocardium, subendocardial or transmural distribution, present in > 1 cardiac phase. A patient was defined as having MFR if at least 2 segments improved by at least one grade each (n = 16) or if the initial echo was normal (n = 3).
Results: Two patients lacked follow-up echo. Eight of the 32 remaining patients lacked MPD on MDCT. Of these 7 (87.5%) showed MFR compared to 12 out of 24 (50%) patients who had a MPD (p = 0.06). Of the 13 patients showing no MFR, 12 (92%) had MPD, whereas of the 19 patients with MFR, 12 (63%) had MPD. Both patients groups (with and without MFR) had similar MPD size, however the enhancement ratios between ischemic myocardium and normal remote myocardium, was significantly lower in patients without MFR (0.55±0.14 vs. 0.67±0.11, p = 0.02). Enhancement ratio > 0.6 gave sensitivity of 84% and specificity of 60% for predicting MFR. Regional dysfunction was found in 6.2±3.2 vs. 3.9±2.9 segments at baseline echo (p = 0.08) and in 5.8±3.2 vs. 2.7±2.4 at follow-up (p = 0.03) for MPD+ and MPD-, respectively. MFR was observed in 1.3±1.6 (MPD+) vs. 3.0±2.5 (MPD-) segments (p = 0.05).
Conclusion: In acute myocardial infarction patients, myocardial perfusion defects on MDCT, and their enhancement ratios are valuable predictors of myocardial functional recovery. Presence of myocardial perfusion defects with reduced enhancement ratios decreases the probability of myocardial functional recovery.
185. MR Angiography of the Coronary Arteries Using Magnetization-Prepared Contrast-Enhanced Breath-Hold Volume-Targeted Imaging (MPCE-VCATS)
Yang C.W.1*; Carr J.C.1; McCarthy R.M.1; Francois C.J.4; Shea S.M.1,2; Deshpande V.S.2,1; Beohar N.3; Meyers S.N.3; Finn J.P.; Li D.1,2; 1. Radiology, Northwestern University Feinberg School of Medicine, Chicago, IL; 2. Biomedical Engineering, Northwestern University Feinberg School of Medicine, Chicago, IL; 3. Cardiology, Northwestern University Feinberg School of Medicine, Chicago, IL; 4. Radiology, Loyola University Medical Center, Maywood, IL; 5. Radiology, UCLA David Geffen School of Medicine, Los Angeles, CA.
Address correspondence to C.W. Yang (cfung{at}alum.mit.edu)
Objective: Coronary artery x-ray angiography (XRA) is currently the gold standard for assessment of coronary artery disease (CAD). A substantial minority of patients referred for XRA have no significant CAD. Magnetization-Prepared Contrast-Enhanced Breath-Hold Volume-Targeted Imaging (MPCE-VCATS) is a new three-dimensional (3D) breath-hold coronary MR angiography technique which combines the use of an extracellular contrast agent together with an inversion recovery magnetization-preparation scheme in order to improve coronary artery signal to noise. The purpose of this study is to evaluate MPCE-VCATS in detecting hemodynamically significant coronary artery stenoses in a patient population with XRA correlation.
Materials and Methods: 19 patients with suspected CAD underwent MPCE-VCATS imaging on a 1.5T Magnetom Sonata (Siemens Medical Systems). All patients had XRA for comparison. Initial breath-hold 3D segmented echo-planar imaging of the heart was acquired and orientations of the proximal right and left coronary arteries were obtained using multiplanar reconstruction (MPR). Two 20ml injections of Gadoteritol (Prohance) were administered to each patient, immediately followed by a breath-hold VCATS scan of the right and left coronary arteries. Imaging parameters were: TR/TE 3.8/1.9 msec; flip angle 20-25 degrees; in-plane resolution 1.4x1.0mm; slice thickness 3mm; 8 partitions/3D slab. An inversion recovery preparation (TI 300 msec) was used to suppress the myocardial signal. 3D data sets were subjected to MPR, maximum intensity projections, and volume-rendering for image interpretation, which was performed independently by 2 observers. The length of vessel visualized and presence/absence of significant (= 50%) stenosis in each coronary artery were noted.
Results: On average, 6.9 cm of the right coronary artery (RCA), 1.8 cm of the left main coronary artery (LCA), and 3.6 cm of the left anterior descending artery (LAD) were visualized. MPCE-VCATS coronary artery imaging showed an overall sensitivity, accuracy, and negative predictive value (NPV) for diagnosing any hemodynamically significant coronary artery disease was 91%, 70%, and 90%, respectively. Sensitivity in the LCA, LAD, and RCA was 100%, 100% and 78%, respectively and NPV was 100%, 100%, and 71%, respectively.
Conclusion: MPCE-VCATS is a promising technique for coronary artery imaging, with a relatively high sensitivity and NPV. The results of the study may indicate the future role of this technique in obviating the need for some patients to undergo XRA.
* Will present paper
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
