Automatic Brain Tumour Segmentation and Biophysics-Guided Survival Prediction

TL;DR

This paper compares advanced CNN models for brain tumour segmentation on mpMRI and introduces a biophysics-guided survival prediction model that outperforms traditional radiomics, achieving second place in a major challenge.

Contribution

It presents a novel biophysics-guided survival prediction model based on ensembled segmentation, improving over existing radiomics methods.

Findings

Ensembled segmentation improves accuracy.

Biophysics-guided model outperforms radiomics.

Achieved second place in MICCAI 2019 BraTS Challenge.

Abstract

Gliomas are the most common malignant brain tumourswith intrinsic heterogeneity. Accurate segmentation of gliomas and theirsub-regions on multi-parametric magnetic resonance images (mpMRI)is of great clinical importance, which defines tumour size, shape andappearance and provides abundant information for preoperative diag-nosis, treatment planning and survival prediction. Recent developmentson deep learning have significantly improved the performance of auto-mated medical image segmentation. In this paper, we compare severalstate-of-the-art convolutional neural network models for brain tumourimage segmentation. Based on the ensembled segmentation, we presenta biophysics-guided prognostic model for patient overall survival predic-tion which outperforms a data-driven radiomics approach. Our methodwon the second place of the MICCAI 2019 BraTS Challenge for theoverall survival prediction.