Groupwise Registration with Physics-Informed Test-Time Adaptation on Multi-parametric Cardiac MRI

TL;DR

This paper introduces a physics-informed deep learning approach with test-time adaptation for accurate groupwise registration of multi-parametric cardiac MRI, improving alignment across different contrast images.

Contribution

It presents a novel physics-informed, test-time adaptive deep learning model for multi-contrast MRI registration, enhancing alignment accuracy across diverse tissue contrasts.

Findings

Improved registration accuracy across multi-contrast MRI images.

Validated on healthy volunteers with various MRI sequences.

Demonstrated robustness to contrast variability.

Abstract

Multiparametric mapping MRI has become a viable tool for myocardial tissue characterization. However, misalignment between multiparametric maps makes pixel-wise analysis challenging. To address this challenge, we developed a generalizable physics-informed deep-learning model using test-time adaptation to enable group image registration across contrast weighted images acquired from multiple physical models (e.g., a T1 mapping model and T2 mapping model). The physics-informed adaptation utilized the synthetic images from specific physics model as registration reference, allows for transductive learning for various tissue contrast. We validated the model in healthy volunteers with various MRI sequences, demonstrating its improvement for multi-modal registration with a wide range of image contrast variability.