PGDM-MRSRGAN: Physics-Guided Degradation Model with an SRGAN Framework for Magnetic Resonance Image Super-Resolution

License: See LICENSE file · Updated 2026-09-24
My rating

This arXiv paper addresses the low signal-to-noise ratio and limited spatial resolution of magnetic resonance imaging, which compromise clinical precision. The proposed approach has two stages: a physics-guided degradation model that models the MRI imaging degradation process, and the Magnetic Resonance Super-Resolution GAN Imaging (MRSRGAN) framework for super-resolution reconstruction. The work targets MRI image reconstruction in the biomedical imaging domain, with applications in low-field MRI.

Tags: