Objective To investigate the feasibility of subtraction coronary computed tomography angiography (Sub-CCTA) for the diagnosis of coronary heart disease in the segment with severe calcification. Methods A retrospective analysis was performed on 27 patients who underwent clinically indicated digital subtraction angiography (CSA) and CCTA using a 320-detector row CT. Compared with the results of DSA, sensitivity, specificity, positive predictive value,negative predictive value and accuracy of Con-CCTA and Sub-CCTA were calculated. The clinical diagnostic accuracy of the two imaging methods was evaluated using the receiver operating characteristic (ROC) curve. The stenosis of coronary segments was divided into four grades ( I,II, III, IV )• Kappa coefficient was used to measure agreement between two imaging methods. Image quality of 4-scale grade scoring method was used and ttest was conducted. Results A total of 52 segments with severe calciicationwere evaluated. The scores of image quality in Con-CCTA and Sub-CCTA were 2. 8±0. 5 and 3. 4±0. 7,respectively. There was significant difference between them (t = 5. 9, P < 0. 05 ). Compared with the resutt of DSA as the golden standard, the Kappa coefficients were 0. 55 and 0. 81 respectively in Con-CCTA and Sub-CCTA for the quantitative evaluation of the severe calcified segments. The sensitivity,specificity,positive predictive value and negative predictive value and accuracy of Con-CCTA were 81.0%, 63.1%, 63.1%, 81.1% and 70.8X; and for Sub-CCTA they were 90. 5X,85. 2%,82. 1 X,92. 0% and 87. 5X respectively. Compared with Con-CCTA,the area under the ROC curve of Con-CCTA and Sub-CCTA were 0. 84 (95%CI:0. 70~ 0.93) and 0.96 (95% CI:0.86- 1.00), respectively, and the difference was statistically significant (P = 0.03). Conclusions Sub-CCTA can improve the diagnostic accuracy of coronary artery stenosis in severe calcified segment. Application of subtraction technique in CCTA can reduce or even eliminate the artifacts caused by severe calcified plaque, and has a good clinical application prospect. [ABSTRACT FROM AUTHOR]