Evaluation and Robustness Assessment of Cardiac Diffusion in Python (CarDpy)

Motivation: Cardiac diffusion tensor imaging (cDTI) can estimate several biomarkers of myocardial remodeling. We evaluated an open-source software package for cDTI processing to enable robust data processing and promote rigorous science. Goal(s): The goal of this study was to evaluate the open-source Cardiac Diffusion in Python (CarDpy) toolbox and identify which post-processing step (POST) most impact improved tensor metric quality. Approach: CarDpy was evaluated at six POSTs. cDTI metrics and their corresponding uncertainties were measured across slices and assessed for significance. Results: cDTI metrics and corresponding uncertainties displayed minimal improvements beyond shot-rejection. Impact: CarDpy is an open-source cardiac diffusion data processing toolbox that enables community development and promotes rigorous science. Each post-processing step was evaluated to estimate the impact on improving the quality of several tensor metrics.