Ensemble of 36 convolutional mesh autoencoders for left-ventricular meshes at end-diastole

Content

This file contains a set of weights and biases for 36 convolutional mesh autoencoder trained on left-ventricular 3D meshes, in turn derived from cardiac magnetic resonance images from the UK Biobank.

The models were built with PyTorch.

Usage

Details on how to use these data to instantiate a model and perform inference can be found on this GitHub repository.