Self-Supervised Multi-Modal Alignment for Whole Body Medical Imaging

TL;DR

This paper presents a self-supervised framework for multi-modal medical image matching, registration, and segmentation transfer using large-scale UK Biobank data, enabling unsupervised cross-modal alignment and segmentation without ground-truth labels.

Contribution

Introduces a contrastive learning approach for multi-modal image matching and registration, enabling unsupervised segmentation transfer across modalities in medical imaging.

Findings

High accuracy in multi-modal scan matching

Unsupervised cross-modal registration achieved

Segmentation transfer without ground-truth labels

Abstract

This paper explores the use of self-supervised deep learning in medical imaging in cases where two scan modalities are available for the same subject. Specifically, we use a large publicly-available dataset of over 20,000 subjects from the UK Biobank with both whole body Dixon technique magnetic resonance (MR) scans and also dual-energy x-ray absorptiometry (DXA) scans. We make three contributions: (i) We introduce a multi-modal image-matching contrastive framework, that is able to learn to match different-modality scans of the same subject with high accuracy. (ii) Without any adaption, we show that the correspondences learnt during this contrastive training step can be used to perform automatic cross-modal scan registration in a completely unsupervised manner. (iii) Finally, we use these registrations to transfer segmentation maps from the DXA scans to the MR scans where they are used to train a network to segment anatomical regions without requiring ground-truth MR examples. To aid further research, our code will be made publicly available.