Semantic Features Aided Multi-Scale Reconstruction of Inter-Modality Magnetic Resonance Images

TL;DR

This paper introduces a deep learning method that reconstructs T2-weighted MR images from T1-weighted images using semantic features, significantly reducing acquisition time while maintaining image quality.

Contribution

A novel encoder-decoder network with semantic features, domain adaptation, and sharp bottleneck modules for fast and accurate inter-modality MR image reconstruction.

Findings

Achieves approximately 1dB PSNR improvement over state-of-the-art methods.

Reconstructs a volume in about 1 second, greatly reducing acquisition time.

Maintains high image quality with negligible artifacts.

Abstract

Long acquisition time (AQT) due to series acquisition of multi-modality MR images (especially T2 weighted images (T2WI) with longer AQT), though beneficial for disease diagnosis, is practically undesirable. We propose a novel deep network based solution to reconstruct T2W images from T1W images (T1WI) using an encoder-decoder architecture. The proposed learning is aided with semantic features by using multi-channel input with intensity values and gradient of image in two orthogonal directions. A reconstruction module (RM) augmenting the network along with a domain adaptation module (DAM) which is an encoder-decoder model built-in with sharp bottleneck module (SBM) is trained via modular training. The proposed network significantly reduces the total AQT with negligible qualitative artifacts and quantitative loss (reconstructs one volume in approximately 1 second). The testing is done on publicly available dataset with real MR images, and the proposed network shows (approximately 1dB) increase in PSNR over SOTA.