Prostate motion modelling using biomechanically-trained deep neural networks on unstructured nodes

TL;DR

This paper introduces a deep neural network approach trained on biomechanical simulations to accurately predict prostate motion during interventions, enabling near real-time inference without subject-specific meshing.

Contribution

It presents a novel method combining biomechanical simulations with deep learning on unstructured point sets, improving speed and generalization in prostate motion prediction.

Findings

Achieved an average prediction error of 0.017 mm.

Demonstrated generalization to new patient data.

Enabled near real-time motion estimation.

Abstract

In this paper, we propose to train deep neural networks with biomechanical simulations, to predict the prostate motion encountered during ultrasound-guided interventions. In this application, unstructured points are sampled from segmented pre-operative MR images to represent the anatomical regions of interest. The point sets are then assigned with point-specific material properties and displacement loads, forming the un-ordered input feature vectors. An adapted PointNet can be trained to predict the nodal displacements, using finite element (FE) simulations as ground-truth data. Furthermore, a versatile bootstrap aggregating mechanism is validated to accommodate the variable number of feature vectors due to different patient geometries, comprised of a training-time bootstrap sampling and a model averaging inference. This results in a fast and accurate approximation to the FE solutions without requiring subject-specific solid meshing. Based on 160,000 nonlinear FE simulations on clinical imaging data from 320 patients, we demonstrate that the trained networks generalise to unstructured point sets sampled directly from holdout patient segmentation, yielding a near real-time inference and an expected error of 0.017 mm in predicted nodal displacement.