Assessment of Pulmonary Arteries Hemodynamics and Its Relationship With Cardiac Remodeling and Myocardial Fibrosis in Athletes With Four-Dimensional Flow MRI

Background: Exercise-induced cardiac remodeling (CR) and myocardial fibrosis (MF) can increase cardiovascular risk in athletes. Early detection of pulmonary arterial hemodynamics parameters among athletes may be beneficial in optimizing the frequency of clinical follow-ups.Purpose: To analyze the hemodynamics of pulmonary arteries and its relationship with CR and MF in athletes using four-dimensional (4D) flow MRI.Study Type: Prospective.Population: One hundred twenty-one athletes (median age, 24 years; mean exercise per week 10 hours, for mean of 5 years) and twenty-one sedentary healthy controls (median age, 25 years; exercise per week <3 hours, irregular pattern).Field Strength/Sequence: True fast imaging with steady state free precession, time-resolved 3D Cartesian phase-contrast, and phase sensitive inversion recovery late gadolinium enhancement sequences at 3.0 T.Assessment: CR was defined as any cardiac parameters exceeding the 99th percentile upper reference limits, encompassing ventricular function, bi-atrium and bi-ventricle diameters, and ventricular wall thickness. MF was visually evaluated by three independent radiologists. 4D flow parameters were assessed in the main, right, and left pulmonary arteries (MPA, RPA, and LPA, respectively) and compared between different groups. Four machine learning (ML) models were developed to differentiate between athletes with and without CR and/or MF.Statistical TestsUnivariate analysis was used to compare groups. Area under the receiver operating characteristic curve (AUC) was used to assess the performance of the ML models.Results: Athletes had significantly higher WSSmax in the MPA, RPA, and LPA than controls. Athletes with CR and/or MF (N = 30) had significantly lower RPmax from MPA to RPA than those without (N = 91). Among the ML models, the gradient boosting machine model had the highest performance, with an AUC of 0.90.Conclusion: The pulmonary arterial hemodynamics parameters could differentiate CR and/or MF in athletes, which may be potential to assist in optimizing frequency of follow-up.