| Part I:Correlation between Myocardial CT Perfusion Imaging and fractional flow reserveObjectives To study the correlation between dynamic CT Myocardial perfusion myocardial blood flow (MBF),myocardial blood volume (MBV), Myocardial Perfusion Reserve (MPR),Myocardial volume Reserve (MVR).myocardial blood flow ratio,myocardial blood volume ratio and fractional flow reserve.Methods A total of twenty-six mini-pigs,body weigh (27.5±2.3Kg),chronic ischemia group 22 mini-pigs,4 mini-pigs in sham group, chronic ischemia group mini-pigs accepted Ameroid narrowing ring on left anterior descending artery, sham group only accepted chest opening,dynamic CT Myocardial perfusion,stress dynamic CT Myocardial perfusion.coronary angiography and fractional flow reserve was performed on the 14th day., Statistical analysis was performed using SPSS 17.0 (SPSS Inc., Chicago, Illinois, USA). FFR as gold standard, the correlation analysis (pearson correlation) between myocardial perfusion index and FFR. Independent samples t-test analysis was performed between myocardial perfusion index and FFR,p< 0.05 were considered statistically significant.Results Four pigs died in chronic ischemia group.two died of ventricular fibrillation after Ameroid narrowing ring placed. one died of ventricular fibrillation after angiography, one died due to heart failure in the rearing process. in the group FFR< 0.80,stress MBF post stenosis was 110.64±15.34 ml/100ml/min vs 132.18±28.78ml/100ml/min in the group FFR>0.80,there were statistical difference (P=0.04). in the group FFR<0.80,stress MBV post stenosis was 9.87±1.83ml/100ml vs 12.1±3.62ml/100ml in the group FFR>0.80, there were no statistical difference (P=0.084). in the group FFR<0.80,MBF reserve was 1.05±0.12 vs 1.31±0.22 in the group FFR≥0.80, there were statistical difference (P=0.002), in the group FFR< 0.80,MBV reserve was 0.82±0.12 vs 1.12±0.38 in the group FFR≥0.80, there were statistical difference (P=0.003). in the group FFR<0.80,MBF ratio was 0.76±0.1 vs 0.91±0.13 in the group FFR≥0.80, there were statistical difference (P=0.007), in the group FFR<0.80,MBV ratio was 0.76±0.1 vs 0.85±0.17 in the group FFR>0.80, there were statistical difference (P=0.007).stress MBF and MBV had moderate correlation with fractional flow reserve(r=0.674 p=0.001; r=0.674 p=0.001), MBF reserve and MBV reserve correlated well with fractional flow reserve(r=0.762 p<0.0001; r=0.803 p< 0.0001), MBF ratio and MBV ratio correlated well with fractional flow reserve(r=0.713 p <0.0001;r=0.761 p<0.0001).Conclusions Myocardial Perfusion Reserve, myocardial volume Reserve,myocardial blood flow ratio, myocardial blood volume ratio correlated well with fractional flow reserve, myocardial blood flow and myocardial blood volume had moderate correlation with fractional flow reserve.PartⅡ :Study of CT dynamic stress myocardial perfusion quantitative evaluation of myocardial ischemiaObjectives:To study the feasibility of stress dynamic myocardial perfusion imaging correlation between dynamic CT Myocardial perfusion myocardial blood flow (MBF),myocardial blood volume (MBV),Myocardial Perfusion Reserve (MPR),Myocardial volume Reserve (MVR),myocardial blood flow ratio,myocardial blood volume ratio for detecting myocardial ischemia in comparison with nuclear MPI.Methods:A total of twenty-six mini-pigs,body weigh (27.5±2.3Kg),chronic ischemia group 22 mini-pigs,4 mini-pigs in sham group, chronic ischemia group mini-pigs accepted Ameroid narrowing ring on left anterior descending artery, sham group only accepted chest opening dynamic CT Myocardial perfusion,stress dynamic CT Myocardial perfusion.coronary angiography and SPECT was performed on the 14th day. Statistical analysis was performed using SPSS 17.0 (SPSS Inc., Chicago, Illinois, USA). SPECT as reference standard,evaluate the diagnostic value of CT Myocardial perfusion index.Results:Four pigs died in chronic ischemia group.two died of ventricular fibrillation after Ameroid narrowing ring placed. one died of ventricular fibrillation after angiography, one died due to heart failure in the rearing process. in the group SPECT negative,stress MBF was 126.59±20.37 ml/100ml/min vs 111.91±10.44 ml/100ml/min in the group SPECT positive,there were statistical difference (p<0.0001). in the group SPECT negative.stress MBV was 12.37±2.63 vs 10.18±1.6ml/100ml in the group SPECT positive,there were statistical difference(p<0.0001). in the group SPECTnegative,MBF reserve was 1.12±0.38 vs 1.06±0.90 in the group SPECT positive, there were statistical difference (p<0.0001), in the group SPECT negative,MBV reserve was 1.07±0.21 vs 0.86±0.080 in the group SPECT positive, there were statistical difference(p<0.0001). in the group SPECT negative,MBF ratio was 0.9±0.1 vs 0.76±0.1 in the group SPECT positive, there were statistical difference (p<0.0001), in the group SPECT negative MBV ratio was0.70±0.92 vs 0.85±0.17 in the group SPECT positive,, there were statistical difference (p< 0.0001). The optimal cut-off value for stress MBF to predict an abnormal SPECT was 111 ml/100ml/min (sensitivity 74.6%, specificity 52.5%),area under ROC curve 0.729(95% interval 0.647-0.810)and The optimal cut-off value for stress MBVto predict an abnormal SPECT was 11.08 ml/100ml (sensitivity 71.1%, specificity 60.6%),area under ROC curve 0.768 (95% interval 0.692-0.844). The optimal cut-off value for MBFreserve to predict an abnormal SPECT was 1.08 (sensitivity 82.7%, specificity 66.7.0%),area under ROC curve 0.809 (95% interval 0.735-0.883). The optimal cut-off value for MBV reserve to predict an abnormal SPECT was 0.91 (sensitivity 82.7%, specificity 66.0%),area under ROC curve 0.836(95% interval 0.773-0.899). The optimal cut-off value for MBF ratio to predict an abnormal SPECT was 0.798(sensitivity 81.6%, specificity 69.7%),area under ROC curve 0.862 (95% interval 0.804-0.919). The optimal cut-off value for MBV ratio to predict an abnormal SPECT was 0.725 (sensitivity80.7%, specificity 54.5%),area under ROC curve 0.827 (95% interval 0.763-0.891).Area under ROC curve to predict an abnormal SPECT for MBF reserve 0.809 was superior to stress MBF 0.7285 (p<0.0001), Area under ROC curve to predict an abnormal SPECT for MBF ratio 0.8614 was superior to stress MBF 0.7285 (p<0.0001), Area under ROC curve to predict an abnormal SPECT for MBV reserve 0.8363was superior to stress MBV 0.7683 (p=0.0248), Area under ROC curve to predict an abnormal SPECT for MBVratio 0.8267 was superior to stress MBV 0.7683 (p=0.0054) Area under ROC curve to predict an abnormal SPECT for MBF reserve 0.809,for MBF ratio 0.8614, there were no statistical difference (p=0.1095), Area under ROC curve to predict an abnormal SPECT for MBV reserve 836 for MBV ratio 0.8267, there were no statistical difference (p=0.6744)Conclusions:Myocardial Perfusion Reserve, myocardial volume Reserve,myocardial blood flow ratio, myocardial blood volume ratio distinguished abnormal SPECT well, Myocardial Perfusion Reserve, myocardial volume Reserve,myocardial blood flow ratio, myocardial blood volume ratio was superior to absolute MBF and MBV,MBF ratio MBV ratio improved the ability to distinguish ischemia without increasing radiation dose.Part Ⅲ:New Morphologic Index characterized by using CT Angiography for prediction of fractional flow reserve Objectives:We aim to test the hypothesis that BARI-score and angiographic index drived from CT correlates well with FFR measurementsMethods:We retrospectively collected sixty-one patients with 72 coronary artery lesions of intermediate severity (50%-80%) who underwent both coronary CT angiography and invasive FFR measurements in our institution. Analysis was only performed in coronary vessels that were ≥2 mm in diameter All observations were blind analyzed by two experienced radiologists (, with 7 years of experience in cardiac imaging; with 6 years of experience in cardiac imaging).All data were transferred to an external workstation (Advantage workstation Version 4.6, GE, Milwaukee, WI, USA) for further analysis. Axial images, cross-sectional views, curved planar reformation (CPRs), and multiplanar reformations, as well as three-dimensional maximum intensity projections (MIPs) were available for evaluation. Quantitative CT measurements were performed using dedicated software.CT measurements included diameter stenosis (DS,%), area stenosis (AS,%), minimal lumen diameter (MLD,mm), minimal lumen area (MLA, mm2), lesion length (LL, mm), CT drived BARI score and the Poiseuille based angiographic index LL/MLD4.Results:The remaining 47 patients (mean age 62.0±10.3; 14 female) with 51 vessels were included for analysis. Patient and vessel characteristics were summarized in The CT BARI score with FFR<0.8 was 0.392±0.076 versus 0.270±0.089 with FFR≥0.8 (p< 0.0001). The optimal cut-off value for BARI to predict an abnormal FFR was 34.1% (sensitivity 81.8%, specificity 78.6%). with an area under curve of 0.852 (95% CI:0.743-0.962) in the ROC analysis.The LL/MLD4 in vessels with FFR<0.8 was 10.29±3.57 versus 5.54±2.40 in vessels with FFR>0.8 (p< 0.0001). The optimal cut-off value for LL/MLD4 to predict an abnormal FFR was 7.31 (sensitivity 82.8%, specificity 79.6%). The area under curve of ROC was 0.898 (CI:0.812-0.983).The LL in vessels with FFR<0.8 was 16.26± 4.86mm versus 11.13±3.50 mm in vessels with FFR≥0.8 (p< 0.001). The optimal cut-off value for LL to predict an abnormal FFR was 12.5mm (sensitivity 72.7%, specificity 75.0%) with an area under curve of 0.808 (CI:0.682-0.935) in the ROC analysis.The AS in vessels with FFR<0.8 was 75.2 ±7.2% versus 61.9±9.8% in vessels with FFR≥0.8 (p< 0.0001). The optimal cut-off value for AS to predict an abnormal FFR was 67.3% (sensitivity 81.8%, specificity 75%) with an area under curve of 0.828 (CI:0.714-0.942) in the ROC analysis. Interobserver agreements for CT BARI score (k= 0.81), LL/MLD4 (k= 0.80), AS (k=0.81), LL (k=0.79), MLA (k=0.80), MLD (k=0.76) were all good. CT BARI score correlates well with angiography BARI score (r=0.908 P= 0.0001)By multivariate regression analysis CT drived BARI score and LL/MLD4,AS,LL were the significant determinants of an abnormal FFR.Conclusion:CT drived BARI score and LL/MLD4,AS,LL were the strong predictors for abnormal FFR and associated with the hemodynamic status of coronary stenoses...
|
PDF Full Text Request