DeepFLASH: An Efficient Network for Learning-based Medical Image Registration

TL;DR

DeepFLASH introduces a low-dimensional, efficient neural network for medical image registration that reduces computational costs while maintaining high accuracy, demonstrated on 2D synthetic and 3D brain MRI data.

Contribution

The paper develops a novel registration network operating in a bandlimited space using complex-valued neural operations, significantly improving efficiency over existing methods.

Findings

Faster registration compared to state-of-the-art methods

Maintains high accuracy in image alignment

Effective on both synthetic and real MRI data

Abstract

This paper presents DeepFLASH, a novel network with efficient training and inference for learning-based medical image registration. In contrast to existing approaches that learn spatial transformations from training data in the high dimensional imaging space, we develop a new registration network entirely in a low dimensional bandlimited space. This dramatically reduces the computational cost and memory footprint of an expensive training and inference. To achieve this goal, we first introduce complex-valued operations and representations of neural architectures that provide key components for learning-based registration models. We then construct an explicit loss function of transformation fields fully characterized in a bandlimited space with much fewer parameterizations. Experimental results show that our method is significantly faster than the state-of-the-art deep learning based image registration methods, while producing equally accurate alignment. We demonstrate our algorithm in two different applications of image registration: 2D synthetic data and 3D real brain magnetic resonance (MR) images. Our code is available at https://github.com/jw4hv/deepflash.