CAR-Net: Unsupervised Co-Attention Guided Registration Network for Joint Registration and Structure Learning

TL;DR

This paper introduces CAR-Net, an unsupervised deep learning model that uses co-attention to improve medical image registration and simultaneously learn structural information without supervision.

Contribution

The paper presents a novel co-attention guided network for unsupervised image registration that enhances accuracy and structural understanding in medical images.

Findings

Higher registration accuracy than state-of-the-art methods

Smoother deformation fields achieved

Provides structural information in an unsupervised manner

Abstract

Image registration is a fundamental building block for various applications in medical image analysis. To better explore the correlation between the fixed and moving images and improve registration performance, we propose a novel deep learning network, Co-Attention guided Registration Network (CAR-Net). CAR-Net employs a co-attention block to learn a new representation of the inputs, which drives the registration of the fixed and moving images. Experiments on UK Biobank cardiac cine-magnetic resonance image data demonstrate that CAR-Net obtains higher registration accuracy and smoother deformation fields than state-of-the-art unsupervised registration methods, while achieving comparable or better registration performance than corresponding weakly-supervised variants. In addition, our approach can provide critical structural information of the input fixed and moving images simultaneously in a completely unsupervised manner.