Unified 3D MRI Representations via Sequence-Invariant Contrastive Learning

TL;DR

This paper introduces a sequence-invariant self-supervised learning framework for 3D MRI that learns anatomy-focused features from multiple contrasts, improving performance and generalization across tasks, protocols, and sites.

Contribution

It proposes a novel contrastive learning method that enforces consistency across MRI contrasts, enabling a single 3D encoder to perform well on various tasks and unseen data.

Findings

Significant improvements in brain and stroke lesion segmentation accuracy.

Enhanced MRI denoising performance, especially with limited data.

Good generalization to unseen sites and protocols.

Abstract

Self-supervised deep learning has accelerated 2D natural image analysis but remains difficult to translate into 3D MRI, where data are scarce and pre-trained 2D backbones cannot capture volumetric context. We present a \emph{sequence-invariant} self-supervised framework leveraging quantitative MRI (qMRI). By simulating multiple MRI contrasts from a single 3D qMRI scan and enforcing consistent representations across these contrasts, we learn anatomy-centric rather than sequence-specific features. The result is a single 3D encoder that excels across tasks and protocols. Experiments on healthy brain segmentation (IXI), stroke lesion segmentation (ARC), and MRI denoising show significant gains over baseline SSL approaches, especially in low-data settings (up to +8.3\% Dice, +4.2 dB PSNR). It also generalises to unseen sites, supporting scalable clinical use. Code and trained models are publicly available at https://github.com/liamchalcroft/contrast-squared