目的：探讨基于放射组学的机器学习（ML）模型在冠状动脉计算机断层血管造影（CCTA）上识别冠状动脉粥样易损病变方面是否优于传统的诊断方法。

材料与方法：回顾性纳入36例合并冠心病（CAD）和终末期心衰的心脏移植受者。收集350个斑块的病理横截面样本，并与患者术前基线CCTA图像进行配准。利用软件从CCTA图像中提取1184个放射性组学特征。通过特征选择和分层五倍交叉验证，我们得到了8个基于放射组学的ML模型用于斑块易损性预测。另外8名接受心脏移植的冠心病独立患者收集196个斑块，用于比较放射组学ML模型与传统CCTA易损征象（至少存在2个高危斑块特征）的诊断准确性。预测模型的性能是通过具有95%置信区间（CI）的受试者工作特征曲线（AUC）下的面积来评估的。

结果：放射组学ML模型的训练组包含200/350 个（57.1%）易损病变，外部验证组包含132/196 个（67.3%）易损斑块。基于8个放射组学特征的基于放射组学的ML模型显示了良好的交叉验证诊断准确性（AUC：0.900±0.033）。在验证组中，基于传统CCTA易损征象的诊断表现中等（AUC：0.656 [95% CI： 0.593-0.718]），而基于放射组学的贝叶斯模型显示更高的诊断能力0.782 [95% CI： 0.710-0.846]（p＜0.0001）。

结论：基于放射组学ML模型在评估冠状动脉斑块易损性方面比传统的CCTA特征具有更好的诊断能力。

Objectives: To explore whether radiomics-based machine learning (ML) models could outperform conventional diagnostic methods at identifying vulnerable lesions on coronary computed tomographic angiography (CCTA).

Material and methods: In this retrospective study, 36 heart transplant recipients with coronary heart disease (CAD) and end-stage heart failure were included. Pathological cross-section samples of 350 plaques were collected and coregistered to patients’ preoperative CCTA images. 1184 radiomic features were extracted from CCTA images. Through feature selection and stratified five-fold cross-validation, we derived eight radiomics-based ML models for lesion vulnerability prediction. An independent set of 196 plaques from another 8 CAD patients who underwent heart transplants was collected to validate radiomics-based ML models’ diagnostic accuracy against conventional CCTA feature-based diagnosis (presence of at least 2 high-risk plaque features). The performance of the prediction models was assessed by the area under the receiver operating characteristic curve (AUC) with 95% Confidence Intervals (CI).

Results: The training group used to develop radiomics-based ML models contained 200/350 (57.1%) vulnerable plaques and the external validation group was composed of 67.3% (132/196) vulnerable plaques. The radiomics-based ML model based on eight radiomic features showed excellent cross-validation diagnostic accuracy (AUC: 0.900 ± 0.033). In the validation group, diagnosis based on conventional CCTA features demonstrated moderate performance (AUC: 0.656 [95% CI: 0.593-0.718]), while the NB model showed higher diagnostic ability 0.782 [95% CI: 0.710-0.846].

Conclusion:  Radiomics-based ML models showed better diagnostic ability than the conventional CCTA features at assessing coronary plaque vulnerability.