Voxel2Hemodynamics: An End-to-End Deep Learning Method for Predicting Coronary Artery Hemodynamics

Local hemodynamic forces play an essential role in determining the functional significance of coronary arterial stenosis and understanding the mechanism of coronary disease progression. Computational fluid dynamics (CFD) has been widely performed to simulate hemodynamics non-invasively from coronary computed tomography angiography (CCTA) images. However, fast and accurate computational analysis is still limited by the complex construction of patient-specific modeling and time-consuming computation. In this work, we proposed an end-to-end deep learning framework, which could predict the coronary artery hemodynamics from CCTA images. The model was trained and evaluated on the hemodynamic data obtained from 3D simulations of ideal synthetic and real datasets. Extensive experiments demonstrated that the predicted hemodynamic distributions by our method agreed well with the CFD-derived results. Quantitatively, the proposed method has the capability of predicting the fractional flow reserve with an average error of 0.5% and 2.5% for the ideal synthetic dataset and real dataset, respectively. This study demonstrates the feasibility and great potential of our end-to-end deep learning method as a fast and accurate approach for hemodynamic analysis. The code can be reached through https://github.com/lullcant/Voxel2Hemodynamic/tree/main .