Chen Meng, Hu Jingcheng, Chen Can, Hao Guangyu, Hu Su, Xu Jialiang, Hu Chunhong
Department of Radiology, The First Affiliated Hospital of Soochow University, Suzhou, PR China.
Institute of Medical Imaging, Soochow University, Suzhou, PR China.
Acta Radiol. 2023 Oct;64(10):2704-2713. doi: 10.1177/02841851231189998. Epub 2023 Aug 21.
Type 2 diabetes mellitus (T2DM) is associated with a markedly increased prevalence of coronary artery disease (CAD). Radiomics features of pericoronary adipose tissue (PCAT) were correlated with inflammation, which may have potential value in the prediction of CAD.
To determine whether radiomics analysis of PCAT captured by plain computed tomography (CT) could predict obstructive CAD in patients with T2DM.
The study included 155 patients with T2DM with suspected CAD between January 2020 and December 2021. Volumes of right coronary artery of 10-50 mm were delineated in the plain CT to extract radiomics features and PCAT CT attenuation (PCATa). Least absolute shrinkage and selection operator was used to select the useful radiomics features to calculate the radiomics score (Rad-score). Univariate and multivariable logistic regression were applied to select independent predictors. The predictive performance was evaluated by the area under the receiver operating characteristics curve (AUC).
Rad-score (per 0.1 increments: odds ratio [OR] = 1.297; < 0.001), coronary artery calcium score (CACS) (OR = 1.003; = 0.037), and sex (OR = 3.245; = 0.026) were identified as independent predictors for obstructive CAD. Rad-score (AUC = 0.835) outperformed CACS (AUC = 0.780), sex (AUC = 0.665), and PCATa (AUC = 0.550) in predicting obstructive CAD ( = 0.017 and 0.003 for Rad-score vs. sex and PCATa, respectively); however, the improvement between Rad-score and CACS had no statistical significance ( = 0.490).
Plain CT-derived Rad-score may be used as a preliminary screening tool for obstructive CAD in patients with T2DM.
2型糖尿病（T2DM）与冠状动脉疾病（CAD）的患病率显著增加相关。冠状动脉周围脂肪组织（PCAT）的放射组学特征与炎症相关，这可能在CAD的预测中具有潜在价值。
确定通过平扫计算机断层扫描（CT）获取的PCAT的放射组学分析能否预测T2DM患者的阻塞性CAD。
本研究纳入了2020年1月至2021年12月期间155例疑似CAD的T2DM患者。在平扫CT上勾勒出10 - 50毫米右冠状动脉的体积，以提取放射组学特征和PCAT CT衰减值（PCATa）。使用最小绝对收缩和选择算子选择有用的放射组学特征以计算放射组学评分（Rad评分）。应用单变量和多变量逻辑回归来选择独立预测因子。通过受试者操作特征曲线下面积（AUC）评估预测性能。
Rad评分（每增加0.1：比值比[OR]=1.297；<0.001）、冠状动脉钙化积分（CACS）（OR = 1.003；=0.037）和性别（OR = 3.245；=0.026）被确定为阻塞性CAD的独立预测因子。在预测阻塞性CAD方面，Rad评分（AUC = 0.835）优于CACS（AUC = 0.780）、性别（AUC = 0.665）和PCATa（AUC = 0.550）（Rad评分与性别和PCATa相比，P分别为0.017和0.003）；然而，Rad评分与CACS之间的改善无统计学意义（=0.490）。
平扫CT得出的Rad评分可作为T2DM患者阻塞性CAD的初步筛查工具。
