Undersampled MRI Reconstruction with Side Information-Guided Normalisation

TL;DR

This paper introduces a novel Side Information-Guided Normalisation (SIGN) module that leverages MRI acquisition side information to enhance undersampled MRI reconstruction quality.

Contribution

The paper proposes a lightweight SIGN module that encodes side information as normalisation parameters, improving existing CNN-based MRI reconstruction methods.

Findings

SIGN improves reconstruction quality across different architectures.

The method is effective on brain and knee MRI datasets.

Significant margin of improvement over baseline models.

Abstract

Magnetic resonance (MR) images exhibit various contrasts and appearances based on factors such as different acquisition protocols, views, manufacturers, scanning parameters, etc. This generally accessible appearance-related side information affects deep learning-based undersampled magnetic resonance imaging (MRI) reconstruction frameworks, but has been overlooked in the majority of current works. In this paper, we investigate the use of such side information as normalisation parameters in a convolutional neural network (CNN) to improve undersampled MRI reconstruction. Specifically, a Side Information-Guided Normalisation (SIGN) module, containing only few layers, is proposed to efficiently encode the side information and output the normalisation parameters. We examine the effectiveness of such a module on two popular reconstruction architectures, D5C5 and OUCR. The experimental results on both brain and knee images under various acceleration rates demonstrate that the proposed method improves on its corresponding baseline architectures with a significant margin.