MrRegNet: Multi-resolution Mask Guided Convolutional Neural Network for Medical Image Registration with Large Deformations

TL;DR

MrRegNet is a multi-resolution, mask-guided deep learning model that improves local and global alignment in medical image registration, especially for large deformations, outperforming traditional and existing deep learning methods.

Contribution

The paper introduces MrRegNet, a novel multi-resolution, mask-guided CNN for medical image registration that effectively handles large deformations and enhances local alignment accuracy.

Findings

Outperforms Demons and VoxelMorph on 3D and 2D MRI datasets

Significantly improves local registration accuracy with segmentation masks

Effective in large deformation scenarios

Abstract

Deformable image registration (alignment) is highly sought after in numerous clinical applications, such as computer aided diagnosis and disease progression analysis. Deep Convolutional Neural Network (DCNN)-based image registration methods have demonstrated advantages in terms of registration accuracy and computational speed. However, while most methods excel at global alignment, they often perform worse in aligning local regions. To address this challenge, this paper proposes a mask-guided encoder-decoder DCNN-based image registration method, named as MrRegNet. This approach employs a multi-resolution encoder for feature extraction and subsequently estimates multi-resolution displacement fields in the decoder to handle the substantial deformation of images. Furthermore, segmentation masks are employed to direct the model's attention toward aligning local regions. The results show that the proposed method outperforms traditional methods like Demons and a well-known deep learning method, VoxelMorph, on a public 3D brain MRI dataset (OASIS) and a local 2D brain MRI dataset with large deformations. Importantly, the image alignment accuracies are significantly improved at local regions guided by segmentation masks. Github link:https://github.com/ruizhe-l/MrRegNet.