Risk prediction model of myocardial fibrosis in patients with coronary heart disease based on CMR-T1 mapping imaging

The objective of this research was to formulate and corroborate a predictive model intended to identify myocardial fibrosis in individuals suffering from coronary artery disease (CAD). This objective was accomplished by integrating radiomic features extracted from cardiac magnetic resonance (CMR) T1 mapping alongside multiple clinical and functional parameters. In this multicenter retrospective analysis, a total of 240 participants diagnosed with CAD who had undergone CMR imaging inclusive of T1 mapping were examined. The cohort was categorized into a training set which included patients from the first center (n = 168), and an independent external validation set, comprising patients from a second center (n = 72). Myocardial fibrosis was identified utilizing the late gadolinium enhancement (LGE) technique. Radiomic features were derived from the T1 mapping images, while pertinent clinical, functional, and tissue characteristics were also gathered. Least absolute shrinkage and selection operator regression method was utilized for feature selection. Subsequently, a multivariate logistic regression model was developed, and its effectiveness was gauged through area under the curve (AUC), decision curve analysis and calibration curves. Key independent predictors identified in the training cohort included left ventricular ejection fraction end-diastolic and end-systolic volume indices, myocardial mass index, post-contrast T1, and extracellular volume fraction (all P < 0.05). The integrated model exhibited strong discriminative performance, with an AUC of 0.955 in the training set and 0.972 in the validation set. Calibration and decision curve analyses further confirmed its clinical applicability. The CMR-based integrated model effectively predicts myocardial fibrosis in CAD patients, offering a non-invasive tool for improved risk stratification and personalized management.