Abstract:
Objective: To evaluate the accuracy of coronary artery calcification score(CACS) by non-gated low-dose chest CT plain scan using artificial intelligence. Methods: A total of 100 cases of patients with coronary CTA scanning were retrospectively analyzed. All patients were selected for ECG gated CT scan and routine non-gated chest CT plain scan. The calcification score of non-gated chest CT plain scan was recorded by Agatston calcification integral software in Siemens post-processing workstation, and the calcification score of non-gated chest CT plain scan was recorded by artificial intelligence analysis software. Two scanning methods were used to obtain Agatston score, including total coronary artery(TOTAL) score, left main artery(LM) score, left Left anterior descending artery(LAD) score, circumflexcoronaryartery(CX) score, and right coronary artery(RCA) score. Paired T-test was used to compare the statistical differences between the two groups. Pearson correlation coefficient was used to analyze the correlation of the two groups' calcification scores, and intra-group correlation coefficient(ICC) was used to test the consistency of risk stratification of the two groups' calcification scores.
P<0.05 was considered statistically significant. Results: there was no statistically significant difference in the Agatston score of LM, CX and RCA between the two groups(
P<0.05), while there was a statistically significant difference in the calcification score of LAD between the two groups(
P<0.05). Pearson analysis found the significant correlation between the two groups in the Agatston score of TOTAL, LM, CX, RCA and LAD. ICC analysis found that the Agatston calcification scores of the two groups showed good consistency in risk stratification, with an intra-group correlation coefficient of 0.938,
P<0.001. Conclusion: The artificial intelligence-based non-gated chest CT plain scan has a high accuracy in the evaluation of coronary artery calcification score, which is in good agreement with the traditional gated examination in terms of risk stratification, and can be used for the screening and assessment of coronary heart disease risk.