The author's main thesis is that Image2Flow, a hybrid neural network, can efficiently segment pulmonary arteries and estimate flow fields from 3D cardiac MRI data. The approach combines image and graph convolutional networks to achieve rapid and accurate results.
Image2Flow enables rapid patient-specific pulmonary artery segmentation and CFD flow field calculation from 3D cardiac MRI data, revolutionizing clinical hemodynamic evaluation.
Image2Flow enables rapid patient-specific pulmonary artery segmentation and CFD flow field calculation from 3D cardiac MRI data.