| Part 1:Preliminary study of the effect of tube voltage on dynamic CT myocardial perfusion measurement of myocardial blood flowObjectivesDecreasing the tube voltage of dynamic CT myocardial perfusion(CTP)reduces the radiation dose,but it is unclear whether tube voltage influences the measurement of myocardial blood flow(MBF).The purpose of this study was to investigate the influence by changing tube voltage on MBF measurements in dynamic CTP.MethodsSeven anesthetized swine(55.8 ± 1.6)kg underwent rest and stress dynamic CTP with tube voltages of 100 kV and 70 kV.The tube current was fixed at 200m.The image noise,signal-to-noise ratio(SNR),contrast-to-noise ratio(CNR),radiation dose.Myocardial perfusion analysis was performed using "CT perfusion 4D" software on images acquired at different tube voltages.The MBF values was obtained for each segment based on the 17-segment American Heart Association(AHA)myocardial standardized segmentation model.Differences in myocardial blood flow(MBF)at the segment level were compared.ResultsThe 70 kV images had a 29.8%higher image noise(27.9±2.4 HU vs.21.5±1.9 HU,P<0.001).The 70 kV images had a 16.5%higher CT attenuation of ventricular septal(77.9±10.6 HU vs.66.8±7.5 HU,P<0.05)and a 71.1%higher CT attenuation of left ventricle cavity(570.2 HU±61.6 HU vs.333.2 HU±39.6 HU,P<0.001)than the 100 kV images,resulting in a higher SNR(20.5±1.6 vs.15.6±1.8,P<0.001)and CNR(17.6±1.5 vs.12.4±1.7,P<0.001).Compared to the use of 100 kV,70 kV yielded an approximately 64.6%radiation dose reduction(348.3±0.7 mGy-cm vs.123.4±0.2 mGy-cm,P<0.001)while generating a comparable MBF values,both at rest(88.3±14.9 ml/100g/min vs.85.6± 17.4 ml/100g/min,P=0.21)and stress(101.4±21.5 ml/100g/min vs.99.6±21.4 ml/100g/min,P=0.58)states at the segment level.ConclusionsDynamic CTP using 70 kV instead of 100 kV does not substantially influence the MBF value but significantly reduces the radiation dose.Reducing tube voltage can be one of the strategies to reduce radiation dose in dynamic CTP.Part 2:Differences in quantification of myocardial blood flow by different CT scanning devicesObjectivesThe threshold of myocardial blood flow(MBF)quantified by dynamic CT myocardial perfusion(CTP)to diagnose myocardial ischemia is highly variable across studies,and whether this difference is related to the scanning device has not been confirmed by relevant studies.Therefore,the purpose of this study was to investigate the difference in MBF values between CT devices.MethodsSeven anesthetized domestic swine(55.8±1.6)kg underwent dynamic CTP with two CT scanning devices(Force CT and Revolution CT,respectively)at rest and stress states.After data acquisition,Revolution CT was used with "CT perfusion 4D" software and Force CT was used with "Volume Perfusion CT Body" software to measure MBF values.The MBF values was obtained for each segment based on the 17-segment American Heart Association(AHA)myocardial standardized segmentation model.Paired Student’s t-test was used to compare the overall differences of MBF values at the level between devices.Coefficient of variation was used to evaluate the degree of dispersion of MBF values at the level of different CT scan devices.One-way ANOVA was used to compare the differences in MBF values in the coronary supply vessel and relative myocardial regions(anterior,lateral,inferior and septum)between devices.ResultsAt the segment level,the quantified MBF values on Revolution CT were higher at rest(81.5±23.6 ml/100g/min vs.68.9±14.5 ml/100g/min,P<0.001)and stress(94.3±25.6 ml/100g/min vs.77.8±18.4 ml/100g/min,P<0.001),The coefficient of variation of Revolution CT was significantly higher than that of Force CT both at rest[(22.7±3.6)%vs.(8.2±2.6)%,P<0.001]and stress[(24.3±4.8)%vs.(7.3±2.0)%,P<0.001].The MBF values of Revolution CT had a higher degree of variability.The mean MBF values quantified by Revolution CT were slightly higher in LAD territories and LCX territories than RCA territories,both at rest and stress.Similarly,the MBF values of the anterior and lateral walls was slightly higher than inferior and septum walls in Revolution CT,both at rest and stress.However,the difference was not significant when compared by groups(all P>0.05).Nevertheless,no such differences were observed in the Force CT among coronary territories and relative myocardial regions,and there was no statistically significant difference when compared by groups(all P>0.05).ConclusionsThe MBF values and its degree of dispersion measured by Revolution CT were higher than those measured by Force CT,and there exist some differences in MBF values measured by Revolution CT in different coronary artery regions and relative myocardial regions.Therefore,MBF values measured by different devices cannot be directly compared in clinical applications.Part 3:Preliminary study on the effect of applying deep learning image reconstruction algorithm on dynamic CT myocardial perfusion measurement of myocardial blood flowObjectivesTo assess the impact on image quality of a new deep learning image reconstruction(DLIR)algorithm in dynamic computed tomography myocardial perfusion imaging(CTP)and whether the algorithm affects the measurement of myocardial blood flow(MBF).MethodsDynamic CTP scanning was performed on five domestic swine(58.6±1.9)kg.The tube voltages were both 100 kV and the low-dose and high-dose scanning tube currents were 150 mA and 300 mA,respectively.The low-dose scan data were reconstructed with filtered back projection(FBP)and three different strengths of DLIR(low,medium,and high).High-dose scan data were reconstructed with FBP only.Objective and subjective(5-point scale)image quality were evaluated,and objective evaluations include image noise,signal-to-noise ratio(SNR),and contrast-to-noise ratio(CNR).The image quality derived from different conditions were compared.The trend test was used to evaluate the trend of image quality with DLIR strengths increase at low dose.The differences of MBF calculated by FBP and the best DLIR CT images at low dose was compared.The Shapiro-Wilk test was used to test the normality of the data,and the paired Student’s t-test was used for the comparison of normal data,and the Wilcoxon rank sum test was used for the comparison of non-normal data.ResultsCompared with the high-dose protocol,the mean effective radiation dose can be reduced by about 47.2%(270.0±0.6 mGy-cm vs.512.0±2.0 mGy-cm,P<0.001)when the low-dose protocol was employed.The image noise gradually decreased and SNR,CNR gradually increased with increasing DLIR strengths(all P for trend<0.001).The image noise(31.7±3.1 HU vs.38.2±1.2 HU,P<0.05),SNR(16.6±2.0 vs.13.8±0.8,P<0.05),CNR(14.5±1.7 vs.11.6±0.9,P<0.05),and subjective score(4.8±0.4 vs.4.2±0.6,P<0.05)derived from low-dose DLIR-high strength were all better than those of high-dose FBP images.There was no statistical difference between low-dose FBP and DLIR-high image on MBF calculation,both at rest(81.3±17.3 ml/100g/min vs.79.9±18.3 ml/100g/min,P=0.27)and stress(99.4±24.9 ml/100g/min vs.100.7±27.3 ml/100g/min,P=0.38)state.ConclusionsDLIR-high strength can improve image quality,facilitating radiation dose savings in CTP without influencing MBF calculation which can be applied in clinical practice. |
PDF Full Text Request