A coronary CT angiography-derived myocardial radiomics model for predicting adverse outcomes in chronic myocardial infarction

Yan Chen, Nan Zhang, Yifeng Gao, Zhen Zhou, Xuelian Gao, Jiayi Liu, Zhifan Gao, Heye Zhang, Zhaoying Wen, Lei Xu*

*此作品的通讯作者

科研成果: 期刊稿件文章同行评审

1 引用 (Scopus)

摘要

Background: The prognostic efficacy of a coronary computed tomography angiography (CCTA)-derived myocardial radiomics model in patients with chronic myocardial infarction (MI) is unclear. Methods: In this retrospective study, a cohort of 236 patients with chronic MI who underwent both CCTA and cardiac magnetic resonance (CMR) examinations within 30 days were enrolled and randomly divided into training and testing datasets at a ratio of 7:3. The clinical endpoints were major adverse cardiovascular events (MACE), defined as all-cause death, myocardial reinfarction and heart failure hospitalization. The entire three-dimensional left ventricular myocardium on CCTA images was segmented as the volume of interest for the extraction of radiomics features. Five models, namely the clinical model, CMR model, clinical+CMR model, CCTA-radiomics model, and clinical+CCTA-radiomics model, were constructed using multivariate Cox regression. The prognostic performances of these models were evaluated through receiver operating characteristic curve analysis and the index of concordance (C-index). Results: Fifty-one (20.16%) patients experienced MACE during a median follow-up of 1439.5 days. The predictive performance of the CCTA-radiomics model surpassed that of the clinical model, CMR model, and clinical+CMR model in both the training (area under the curve (AUC) of 0.904 vs. 0.691, 0.764, 0.785; C-index of 0.88 vs. 0.71, 0.75, 0.76, all p values <0.001) and testing (AUC of 0.893 vs. 0.704, 0.851, 0.888; C-index of 0.86 vs. 0.73, 0.85, 0.85, all p values <0.05) datasets. Conclusions: The CCTA-based myocardial radiomics model is a valuable tool for predicting adverse outcomes in chronic MI, providing incremental value to conventional clinical and CMR parameters.

源语言英语
文章编号132265
期刊International Journal of Cardiology
411
DOI
出版状态已出版 - 15 9月 2024
已对外发布

指纹

探究 'A coronary CT angiography-derived myocardial radiomics model for predicting adverse outcomes in chronic myocardial infarction' 的科研主题。它们共同构成独一无二的指纹。

引用此