Learning Two-factor Representation for Magnetic Resonance Image Super-resolution

TL;DR

This work factors intensity signals into a linear combination of learnable basis and coefficient factors, enabling efficient continuous volumetric representation from low-resolution MR image, and introduces a coordinate-based encoding to capture structural relationships between sparse voxels, facilitating smooth completion in unobserved regions.

Abstract

Magnetic Resonance Imaging (MRI) requires a trade-off between resolution, signal-to-noise ratio, and scan time, making high-resolution (HR) acquisition challenging. Therefore, super-resolution for MR image is a feasible solution. However, most existing methods face challenges in accurately learning a continuous volumetric representation from low-resolution image or require HR image for supervision. To solve these challenges, we propose a novel method for MR image super-resolution based on two-factor representation. Specifically, we factorize intensity signals into a linear combination of learnable basis and coefficient factors, enabling efficient continuous volumetric representation from low-resolution MR image. Besides, we introduce a coordinate-based encoding to capture structural relationships between sparse voxels, facilitating smooth completion in unobserved regions. Experiments on BraTS 2019 and MSSEG 2016 datasets demonstrate that our method achieves state-of-the-art performance, providing superior visual fidelity and robustness, particularly in large up-sampling scale MR image super-resolution.

Figures (2)

• Figure 1: Overview of proposed method. The upper part shows the training process, and the lower part shows the testing process.
• Figure 2: Comparison of qualitative results on BraTS 2019 BraTS and MSSEG 2016 MSSEG datasets. We set up-sampling scale to 4.