Comprehensive assessment of cardiac structure and function in HFpEF like models using MRI and machine learning-based quantification

Motivation: Early detection of cardiovascular diseases is crucial, but conventional methods often miss subtle changes. Cardiac MRI offers high sensitivity, but traditional analysis methods are time-consuming and subject to observer bias. Goal(s): To optimize a preclinical cardiac MRI protocol for fast acquisition and high sensitivity to detect early cardiac changes in mouse models of obesity and HFpEF. Approach: We refined cardiac MRI acquisition parameters and implemented a machine learning algorithm for data analysis. Results: Automated analysis reduced variability and saved time. Early vascular remodeling despite preserved cardiac function was observed in obese mice. In HFpEF model, we identified subtle diastolic dysfunction and vascular remodeling. Impact: This optimized protocol enables sensitive and efficient detection of subtle cardiac changes, providing a valuable tool for preclinical research and advancing our understanding of cardiovascular disease progression, particularly in the context of HFpEF.