Qualitative and Quantitative Diagnostic Performance of 320-Slice Computed Tomography for Detecting Coronary Artery Disease with Respect to Atherosclerotic Plaque Characteristics

Suhua Li,Jinlai Liu,Long Peng,Ruimin Dong,Huilan Wu,Chenlin Wang,Qiongqiong Ni,Yanting Luo,Jieming Zhu,Lin Chen
DOI: https://doi.org/10.3760/cma.j.issn.0376-2491.2014.39.009
2014-01-01
Abstract:Objective To investigate qualitatively and quantitatively the diagnostic performance of 320-slice CT for detection of coronary artery disease with respect to different atherosclerotic plaque characteristics.Methods A retrospective search was performed for inpatients underwent both coronary CT and further coronary angiography (CAG) from December 1,2008 to December 31,2012.The diagnostic performance of 320-slice CTA for detecting significant stenosis (≥ 50% diameter) with respect to atherosclerotic plaque characteristics were analyzed by calculating sensitivity,specificity,positive predictive value (PPV),negative predictive value (NPV),accuracy,kappa index (κ),and area under the receiver operating characteristic curve (AUC).Chi-square test was used to evaluate whether there were significant differences of the true-case frequency (true positive + true negative) and false-case frequency (false positive + false negative) among groups.Bland-Altman analysis was used to determine limits of agreement between CTA and CAG.Results A total of 454 patients and 6 779 segments were analyzed.Diagnostic accuracy was higher in non-calcified segments; whereas they decreased in the presence of both mildmoderately and heavily calcified plaques.Excellent agreement (κ =0.810) between CT and CAG was observed for non-calcified segments,while good agreement was observed for both mild-moderately (κ =0.701) and heavily calcified segments (κ =0.750).Both mild-moderate (P =0.000) and heavy (P =0.000) calcification decreased the true-case frequency and increased the false-case frequency when compared to non-calcification.There were no significant underestimation or overestimation for non-calcified (P =0.087) and mild-moderately calcified (P =0.704) segments,while there was significant overestimation for heavily calcified segments (P =0.001).Conclusions Great qualitative and quantitative diagnostic performances of 320-slice CT were observed in non-calcified coronary segments.However,qualitative diagnostic performance decreased in both mild-moderately and heavily calcified segments,and quantitative overestimation were observed in heavily calcified segments.
What problem does this paper attempt to address?