Influence Factors On Accuracy Of CT Attenuation Measurement Of Coronary Artery Non-calcified Plaque: Preliminary Study

【Objective】Multislice computed tomography (MSCT) coronary angiography has showedgood potential in the detection, quantification, and characterization of coronary arteryplaques. However, there are many factors affecting on the accuracy of CT attenuationmeasurement. The purpose of this study was to assess the effect of contrast mediumconcentration, slice thickness, and window settings on CT attenuation measurement,and provide the evidence for the standardization of CT attenuation measurement.【Materials and Methods】Select 20 adult cadaver heart specimens (Male, 13; Female, 7; 50～70 years old,average 57.6). Four saline solutions with decreasing dilution of contrast medium(300rag I/mL) were used: 1:50, 1:40, 1:30, and 1:20. The solutions wereinjected into the lumen of coronary artery specimens, and a MSCT scan (GELightSpeed Pro 16) was performed. Observing the CT images to find the non-calcifiedplaques with the luminal stenosis of 30～75％, and measure the CT attenuation withregion of interest (ROI) under four contrast medium concentrations, two slice thickness, and two window settings. The CT attenuation of every part was comparedby ANOVA or SNK-q, and the size of test was 0.05. The correlation between contrastmedium concentrations, slice thickness, window settings, and CT attenuation wereanalyzed. The CT attenuation as independent variable and contrast mediumconcentrations (CT attenuation of lumen), slice thickness, window settings asindependent variables were analyzed by multiple linear regression analysis.【Results】In the 20 adult cadaver heart specimens, 16 hearts had non-calcified plaques.There were 41 non-calcified plaques in all, and 24 with the luminal stenosis of30～75％.1. The effect of contrast medium concentration (CT attenuation of lumen) onCT attenuation measurement.Contrast medium concentration (CT attenuation of lumen) had significanteffect on CT attenuation measurement, and there was positive correlationbetween them (r=0.450, P=0.000). With the concentration increasing, CTattenuation showed increasing tendency. The increasing range was 6.13～32.08HU.2. The effect of slice thickness on CT attenuation measurement.Slice thickness showed no significant correlation with CT attenuation (r=0.070, P=0.102).3. The effect of window settings on CT attenuation measurement.Window settings had significant effect on CT attenuation measurement, andthere was correlation between them (r=0.351, P=0.000).4. Multiple linear regression analysis between CT attenuation and effectfactors. There was linear regression relationship between CT attenuation of lumen,window settings, and CT attenuation (P＜0.001). The equation of multipleregression was: CT attenuation of plaque=0.067×(CT attenuation of lumen)+10.306×(window settings)+7.311.【Conclusion】1. Contrast medium concentration (CT attenuation of lumen) had significanteffect on CT attenuation measurement. With the concentration increasing,CT attenuation of plaque showed increasing tendency.2. Slice thickness showed no significant effect on CT attenuation measurementof plaque.3. Window settings had significant effect on CT attenuation measurement ofplaque.4. The equation of multiple regression was: CT attenuation of plaque=0.067×(CT attenuation of lumen)+10.306×(window settings)+7.311.【Innovative point】1. To study the influence factors of CT attenuation measurement of plaqueswith cadaver heart specimens.2. To achieve the multiple regression equation of CT attenuation of plaque, CTattenuation of lumen, and window settings.