Contrastive Anatomy-Contrast Disentanglement: A Domain-General MRI Harmonization Method

2025-09-08

AI Summary

MRI contrast inconsistency across multi-center scanners—caused by heterogeneous acquisition protocols—severely compromises comparability and reproducibility in multi-site and longitudinal studies. To address this, we propose an anatomy–contrast disentangled conditional diffusion autoencoder framework, integrating contrastive learning with domain-invariant contrast enhancement, enabling cross-scanner brain MRI synthesis and generalization without subject re-scanning or model fine-tuning. Our method standardizes contrast while preserving fine anatomical details. Experiments demonstrate a 7% PSNR improvement on the Traveling Subjects dataset and an 18% performance gain in age regression on unseen domains—substantially outperforming existing baselines. The core contribution lies in the first unified paradigm that jointly leverages disentangled representation learning, contrastive learning, and diffusion-based generation for unpaired, fine-tuning-free MRI contrast normalization across heterogeneous scanner domains.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Magnetic resonance imaging (MRI) is an invaluable tool for clinical and research applications. Yet, variations in scanners and acquisition parameters cause inconsistencies in image contrast, hindering data comparability and reproducibility across datasets and clinical studies. Existing scanner harmonization methods, designed to address this challenge, face limitations, such as requiring traveling subjects or struggling to generalize to unseen domains. We propose a novel approach using a conditioned diffusion autoencoder with a contrastive loss and domain-agnostic contrast augmentation to harmonize MR images across scanners while preserving subject-specific anatomy. Our method enables brain MRI synthesis from a single reference image. It outperforms baseline techniques, achieving a +7% PSNR improvement on a traveling subjects dataset and +18% improvement on age regression in unseen. Our model provides robust, effective harmonization of brain MRIs to target scanners without requiring fine-tuning. This advancement promises to enhance comparability, reproducibility, and generalizability in multi-site and longitudinal clinical studies, ultimately contributing to improved healthcare outcomes.

Problem

• Harmonizing MRI contrasts across different scanners
• Eliminating data inconsistencies from acquisition variations
• Preserving anatomical details while standardizing image contrast

Innovation

• Conditioned diffusion autoencoder with contrastive loss
• Domain-agnostic contrast augmentation for harmonization
• Single reference image synthesis across scanners