Self-supervised iRegNet for the Registration of Longitudinal Brain MRI of Diffuse Glioma Patients

TL;DR

This paper introduces an enhanced self-supervised version of iRegNet for accurate longitudinal brain MRI registration in glioma patients, achieving significant error reduction and competitive challenge results.

Contribution

The paper presents a novel self-supervised extension to iRegNet, improving registration accuracy for pathological brain MRI scans in a challenging longitudinal setting.

Findings

Reduced median MAE from 8.20 mm to 3.51 mm on training data

Achieved an MAE of 2.93 mm on validation data

Ranked 5th in the BraTS-Reg 2022 challenge

Abstract

Reliable and accurate registration of patient-specific brain magnetic resonance imaging (MRI) scans containing pathologies is challenging due to tissue appearance changes. This paper describes our contribution to the Registration of the longitudinal brain MRI task of the Brain Tumor Sequence Registration Challenge 2022 (BraTS-Reg 2022). We developed an enhanced unsupervised learning-based method that extends the iRegNet. In particular, incorporating an unsupervised learning-based paradigm as well as several minor modifications to the network pipeline, allows the enhanced iRegNet method to achieve respectable results. Experimental findings show that the enhanced self-supervised model is able to improve the initial mean median registration absolute error (MAE) from 8.20 (7.62) mm to the lowest value of 3.51 (3.50) for the training set while achieving an MAE of 2.93 (1.63) mm for the validation set. Additional qualitative validation of this study was conducted through overlaying pre-post MRI pairs before and after the de-formable registration. The proposed method scored 5th place during the testing phase of the MICCAI BraTS-Reg 2022 challenge. The docker image to reproduce our BraTS-Reg submission results will be publicly available.