MHAN: Multi-Stage Hybrid Attention Network for MRI reconstruction and super-resolution

Wanliang Wang; Haoxin Shen; Jiacheng Chen; Fangsen Xing

doi:10.1016/j.compbiomed.2023.107181

MHAN: Multi-Stage Hybrid Attention Network for MRI reconstruction and super-resolution

Comput Biol Med. 2023 Sep:163:107181. doi: 10.1016/j.compbiomed.2023.107181. Epub 2023 Jun 16.

Authors

Wanliang Wang¹, Haoxin Shen², Jiacheng Chen³, Fangsen Xing⁴

Affiliations

¹ College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou, 310023, China. Electronic address: zjutwwl@zjut.edu.cn.
² College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou, 310023, China. Electronic address: 2112112071@zjut.edu.cn.
³ College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou, 310023, China. Electronic address: 15968812143@163.com.
⁴ College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou, 310023, China. Electronic address: xfangsen@163.com.

PMID: 37352637
DOI: 10.1016/j.compbiomed.2023.107181

Abstract

High-quality magnetic resonance imaging (MRI) affords clear body tissue structure for reliable diagnosing. However, there is a principal problem of the trade-off between acquisition speed and image quality. Image reconstruction and super-resolution are crucial techniques to solve these problems. In the main field of MR image restoration, most researchers mainly focus on only one of these aspects, namely reconstruction or super-resolution. In this paper, we propose an efficient model called Multi-Stage Hybrid Attention Network (MHAN) that performs the multi-task of recovering high-resolution (HR) MR images from low-resolution (LR) under-sampled measurements. Our model is highlighted by three major modules: (i) an Amplified Spatial Attention Block (ASAB) capable of enhancing the differences in spatial information, (ii) a Self-Attention Block with a Data-Consistency Layer (DC-SAB), which can improve the accuracy of the extracted feature information, (iii) an Adaptive Local Residual Attention Block (ALRAB) that focuses on both spatial and channel information. MHAN employs an encoder-decoder architecture to deeply extract contextual information and a pipeline to provide spatial accuracy. Compared with the recent multi-task model T2Net, our MHAN improves by 2.759 dB in PSNR and 0.026 in SSIM with scaling factor ×2 and acceleration factor 4× on T2 modality.

Keywords: Hybrid attention mechanism; Magnetic resonance; Reconstruction; Super-resolution.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Image Processing, Computer-Assisted* / methods
Magnetic Resonance Imaging*