CAR: Contrast-Agnostic Deformable Medical Image Registration with Contrast-Invariant Latent Regularization

TL;DR

This paper introduces CAR, a contrast-agnostic deformable image registration framework that uses contrast augmentation and latent regularization to achieve accurate registration across various imaging contrasts without prior exposure.

Contribution

The work presents a novel contrast-agnostic registration method with contrast augmentation and contrast-invariant latent regularization, enabling generalization to unseen contrasts.

Findings

CAR outperforms baseline methods in registration accuracy.

CAR demonstrates superior generalization to unseen imaging contrasts.

The proposed approach is faster than traditional iterative methods.

Abstract

Multi-contrast image registration is a challenging task due to the complex intensity relationships between different imaging contrasts. Conventional image registration methods are typically based on iterative optimizations for each input image pair, which is time-consuming and sensitive to contrast variations. While learning-based approaches are much faster during the inference stage, due to generalizability issues, they typically can only be applied to the fixed contrasts observed during the training stage. In this work, we propose a novel contrast-agnostic deformable image registration framework that can be generalized to arbitrary contrast images, without observing them during training. Particularly, we propose a random convolution-based contrast augmentation scheme, which simulates arbitrary contrasts of images over a single image contrast while preserving their inherent structural information. To ensure that the network can learn contrast-invariant representations for facilitating contrast-agnostic registration, we further introduce contrast-invariant latent regularization (CLR) that regularizes representation in latent space through a contrast invariance loss. Experiments show that CAR outperforms the baseline approaches regarding registration accuracy and also possesses better generalization ability to unseen imaging contrasts. Code is available at \url{https://github.com/Yinsong0510/CAR}.