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

TL;DR

This paper introduces a novel two-factor representation method for MRI super-resolution that efficiently models continuous volumetric data and captures structural relationships, achieving state-of-the-art results in large up-sampling scales.

Contribution

It proposes a new two-factor representation approach with coordinate-based encoding for MRI super-resolution, addressing limitations of existing methods in continuous modeling and supervision.

Findings

Achieves state-of-the-art performance on BraTS 2019 and MSSEG 2016 datasets.

Provides superior visual fidelity and robustness in large up-sampling scales.

Effectively models continuous volumetric MRI data from low-resolution images.

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.