Test-Time Training for Deformable Multi-Scale Image Registration

Registration is a fundamental task in medical robotics and is often a crucial step for many downstream tasks such as motion analysis, intra-operative tracking and image segmentation. Popular registration methods such as ANTs and NiftyReg optimize objective functions for each pair of images from scratch, which are time-consuming for 3D and sequential images with complex deformations. Recently, deep learning-based registration approaches such as VoxelMorph have been emerging and achieve competitive performance. In this work, we construct a test-time training for deep deformable image registration to improve the generalization ability of conventional learning-based registration model. We design multi-scale deep networks to consecutively model the residual deformations, which is effective for high variational deformations. Extensive experiments validate the effectiveness of multi-scale deep registration with test-time training based on Dice coefficient for image segmentation and mean square error (MSE), normalized local cross-correlation (NLCC) for tissue dense tracking tasks.

I. Introduction

Image registration tries to establish the correspondence between organs, tissues, landmarks, edges, or surfaces in different images and it is critical to many clinical tasks such as tumor growth monitoring and surgical robotics [1]. Manual image registration is time-consuming, laborious and lacks reproducibility which causes clinical disadvantage potentially. Thus, automated registration is desired in many clinical practices. Generally, registration can be necessary to analyze motion from videos, auto-segment organs given atlases, and align a pair of images from different modalities, acquired at different times, from different viewpoints or even from different patients. Thus designing a robust image registration can be challenging due to the high variability of scenarios.