T1 Mapping–derived Predictors of Cardiac Remodeling and Fibrosis in Athletes using Advanced Machine Learning...

This study aimed to predict the occurrence of cardiac remodeling and /or myocardial fibrosis with machine learning based on T1 mapping of cardiovascular magnetic resonance in athletes.

A total of 104 athletes and 20 healthy sedentary controls underwent a 3.0T cardiovascular magnetic resonance scan. Cardiac function parameters, T1, and extracellular volume values of 16 segments for the left ventricle were measured, respectively. These parameters were separately compared between athletes and controls, as well as between the positive and negative athlete groups. Four machine learning models were constructed for the prediction of cardiac remodeling and /or myocardial fibrosis.

The most effective model was Gradient Boosting Machine, with an AUC value of 0.899, accuracy of 82.7%, sensitivity of 90.0%, and specificity of 81.0%. The top three important factors were the native T1 value of segment 10, the extracellular volume value of segment 3, and body surface area.

The direct reason for the increased native T1 value and ECV observed in our study was CR for athletes, which may reveal the relationship between CR and MF. The so-called physiological CR of athletes has certain potential risks. Furthermore, the limitations of this study are that only male athletes were included, and the sample size of the control group was small. This study was a single-center study, and there were selection biases.

Native T1 and extracellular volume values increased in athletes with cardiac remodeling, which may reveal the relationship between cardiac remodeling and myocardial fibrosis. Early cardiac magnetic resonance imaging is conducted to monitor the myocardial native T1 and ECV values of athletes, assess their risk levels, and guide their subsequent surge planning to reduce the occurrence of adverse cardiovascular events.