Generative 3D Cardiac Shape Modelling for In-Silico Trials

TL;DR

This paper introduces a deep learning approach for modeling and generating synthetic 3D aortic shapes using neural signed distance fields, enabling realistic in-silico trials and shape variability analysis.

Contribution

It presents a novel neural signed distance field method conditioned on trainable embeddings for high-fidelity 3D shape generation of aortic anatomies.

Findings

High-fidelity shape representation

Ability to generate realistic synthetic aortic shapes

Potential for use in in-silico clinical trials

Abstract

We propose a deep learning method to model and generate synthetic aortic shapes based on representing shapes as the zero-level set of a neural signed distance field, conditioned by a family of trainable embedding vectors with encode the geometric features of each shape. The network is trained on a dataset of aortic root meshes reconstructed from CT images by making the neural field vanish on sampled surface points and enforcing its spatial gradient to have unit norm. Empirical results show that our model can represent aortic shapes with high fidelity. Moreover, by sampling from the learned embedding vectors, we can generate novel shapes that resemble real patient anatomies, which can be used for in-silico trials.