A radiomic nomogram for prediction of major adverse cardiac events in ST-segment elevation myocardial infarction.

Abstract

This study was conducted to establish and validate a non-contrast T1 map-based radiomic nomogram for predicting major adverse cardiac events (MACEs) in patients with acute ST-segment elevation myocardial infarction (STEMI) undergoing percutaneous coronary intervention (PCI). This retrospective study included 157 consecutive patients (training sets, 109 patients; test sets, 48 patients) with acute STEMI undergoing PCI. An open-source radiomics software was used to segment the myocardium on the non-contrast T1 mapping and extract features. A radiomic signature was constructed to predict MACEs using the least absolute shrinkage and selection operator method. The performance of the radiomic nomogram for predicting MACEs in both the training and test sets was evaluated by its discrimination, calibration, and clinical usefulness. The radiomic signature showed a good prognostic ability in the training sets with an AUC of 0.94 (95% CI, 0.86 to 1.00) and F1 score of 0.71, which was confirmed in the test sets with an AUC of 0.90 (95% CI, 0.74 to 1.00) and F1 score of 0.62. The nomogram consisting of the radiomic scores and cardiac troponin I showed good discrimination ability in the training and test sets with AUCs of 0.96 (95% CI, 0.91 to 1.00; F1 score, 0.71) and 0.94 (95% CI, 0.83 to 1.00; F1 score, 0.70), respectively. The non-contrast T1 map-based radiomic nomogram is a useful tool for the prediction of MACEs in patients with acute STEMI undergoing PCI that can assist clinicians for optimised risk stratification of individual patients. • Radiomic signature improved MACE prediction in acute STEMI patients. • T1 mapping-derived radiomic signature outperformed conventional cardiac MRI parameters in predicting MACEs in acute STEMI patients. • The non-contrast T1 mapping-based radiomic nomogram can be used for prediction of MACEs and improvement of risk stratification in acute STEMI.