Deform-GAN:An Unsupervised Learning Model for Deformable Registration

TL;DR

Deform-GAN introduces an unsupervised deep learning approach for 3D medical image deformable registration, utilizing gradient and adversarial losses to improve accuracy across modalities without manual labels.

Contribution

This is the first deep learning registration method to incorporate gradient loss and adversarial training for robust multi-modal alignment.

Findings

Outperforms existing methods in accuracy and speed.

Effectively handles noise, blur, and non-functional intensity relations.

Does not require ground-truth or manual labels during training.

Abstract

Deformable registration is one of the most challenging task in the field of medical image analysis, especially for the alignment between different sequences and modalities. In this paper, a non-rigid registration method is proposed for 3D medical images leveraging unsupervised learning. To the best of our knowledge, this is the first attempt to introduce gradient loss into deep-learning-based registration. The proposed gradient loss is robust across sequences and modals for large deformation. Besides, adversarial learning approach is used to transfer multi-modal similarity to mono-modal similarity and improve the precision. Neither ground-truth nor manual labeling is required during training. We evaluated our network on a 3D brain registration task comprehensively. The experiments demonstrate that the proposed method can cope with the data which has non-functional intensity relations, noise and blur. Our approach outperforms other methods especially in accuracy and speed.