MRI Super-Resolution with Deep Learning: A Comprehensive Survey
Mohammad Khateri, Serge Vasylechko, Morteza Ghahremani, Liam Timms, Deniz Kocanaogullari, Simon K. Warfield, Camilo Jaimes, Davood Karimi, Alejandra Sierra, Jussi Tohka, Sila Kurugol, Onur Afacan
2025-12-01
Summary
This paper is a comprehensive overview of a technique called 'super-resolution' MRI, which aims to create highly detailed MRI images without needing expensive equipment or extremely long scan times.
What's the problem?
Getting really clear, high-resolution MRI images is usually expensive and takes a lot of time. It's a balancing act between image quality, how long the patient is in the scanner, and the cost of the technology. Traditional methods have limitations, making it difficult to get the best possible images efficiently.
What's the solution?
The paper focuses on how 'deep learning,' a type of artificial intelligence, can be used to solve this problem. Researchers are developing computer programs that can take lower-quality MRI scans and intelligently 'fill in the gaps' to create images that *look* like high-resolution scans. The paper breaks down the different ways these programs are designed, how they learn, and how well they perform, looking at it from multiple angles like computer science, medical imaging, and physics.
Why it matters?
This research is important because it could make MRI scans faster, cheaper, and more accessible. Better image quality can lead to more accurate diagnoses and improved patient care, all without requiring hospitals to invest in significantly more expensive hardware. The paper also provides resources for other researchers to build on this work, accelerating progress in the field.
Abstract
High-resolution (HR) magnetic resonance imaging (MRI) is crucial for many clinical and research applications. However, achieving it remains costly and constrained by technical trade-offs and experimental limitations. Super-resolution (SR) presents a promising computational approach to overcome these challenges by generating HR images from more affordable low-resolution (LR) scans, potentially improving diagnostic accuracy and efficiency without requiring additional hardware. This survey reviews recent advances in MRI SR techniques, with a focus on deep learning (DL) approaches. It examines DL-based MRI SR methods from the perspectives of computer vision, computational imaging, inverse problems, and MR physics, covering theoretical foundations, architectural designs, learning strategies, benchmark datasets, and performance metrics. We propose a systematic taxonomy to categorize these methods and present an in-depth study of both established and emerging SR techniques applicable to MRI, considering unique challenges in clinical and research contexts. We also highlight open challenges and directions that the community needs to address. Additionally, we provide a collection of essential open-access resources, tools, and tutorials, available on our GitHub: https://github.com/mkhateri/Awesome-MRI-Super-Resolution. IEEE keywords: MRI, Super-Resolution, Deep Learning, Computational Imaging, Inverse Problem, Survey.