A Deep Discontinuity-Preserving Image Registration Network

TL;DR

This paper introduces a deep learning-based image registration network that preserves tissue discontinuities, improving accuracy and realism in medical image registration, especially at tissue interfaces where traditional methods struggle.

Contribution

The proposed weakly-supervised DDIR network effectively models discontinuities in deformation fields, addressing limitations of smoothness assumptions in existing methods.

Findings

Achieves higher registration accuracy on cardiac MR images.

Predicts more realistic deformation fields at tissue interfaces.

Outperforms state-of-the-art registration approaches.

Abstract

Image registration aims to establish spatial correspondence across pairs, or groups of images, and is a cornerstone of medical image computing and computer-assisted-interventions. Currently, most deep learning-based registration methods assume that the desired deformation fields are globally smooth and continuous, which is not always valid for real-world scenarios, especially in medical image registration (e.g. cardiac imaging and abdominal imaging). Such a global constraint can lead to artefacts and increased errors at discontinuous tissue interfaces. To tackle this issue, we propose a weakly-supervised Deep Discontinuity-preserving Image Registration network (DDIR), to obtain better registration performance and realistic deformation fields. We demonstrate that our method achieves significant improvements in registration accuracy and predicts more realistic deformations, in registration experiments on cardiac magnetic resonance (MR) images from UK Biobank Imaging Study (UKBB), than state-of-the-art approaches.