Recursive Deformable Image Registration Network with Mutual Attention

TL;DR

This paper introduces a recursive deformable image registration network with mutual attention, significantly improving accuracy in medical image registration tasks across lung and abdominal CT datasets.

Contribution

It proposes a novel recursive network architecture with mutual attention to address receptive field limitations in multi-stage registration methods.

Findings

Achieves highest accuracy in lung CT registration with a Dice score of 92%.

Attains competitive results in abdominal CT with a Dice score of 55%.

Adding three recursive networks yields state-of-the-art performance without much inference time increase.

Abstract

Deformable image registration, estimating the spatial transformation between different images, is an important task in medical imaging. Many previous studies have used learning-based methods for multi-stage registration to perform 3D image registration to improve performance. The performance of the multi-stage approach, however, is limited by the size of the receptive field where complex motion does not occur at a single spatial scale. We propose a new registration network combining recursive network architecture and mutual attention mechanism to overcome these limitations. Compared with the state-of-the-art deep learning methods, our network based on the recursive structure achieves the highest accuracy in lung Computed Tomography (CT) data set (Dice score of 92\% and average surface distance of 3.8mm for lungs) and one of the most accurate results in abdominal CT data set with 9 organs of various sizes (Dice score of 55\% and average surface distance of 7.8mm). We also showed that adding 3 recursive networks is sufficient to achieve the state-of-the-art results without a significant increase in the inference time.