An Application of Generative Adversarial Networks for Super Resolution Medical Imaging

TL;DR

This paper explores using SRGAN to enhance low resolution prostate MR images, improving visual quality and resolution while reducing scan time and patient discomfort, compared to other super resolution models.

Contribution

It applies SRGAN to medical imaging, demonstrating its effectiveness in improving visual quality of super-resolved MR images over other models.

Findings

SRGAN produces visually superior images despite lower PSNR and SSIM.

Super resolution factors of 4 and 8 are achieved.

SRGAN outperforms SRCNN, SRResNet, and Sparse Representation in visual similarity.

Abstract

Acquiring High Resolution (HR) Magnetic Resonance (MR) images requires the patient to remain still for long periods of time, which causes patient discomfort and increases the probability of motion induced image artifacts. A possible solution is to acquire low resolution (LR) images and to process them with the Super Resolution Generative Adversarial Network (SRGAN) to create an HR version. Acquiring LR images requires a lower scan time than acquiring HR images, which allows for higher patient comfort and scanner throughput. This work applies SRGAN to MR images of the prostate to improve the in-plane resolution by factors of 4 and 8. The term 'super resolution' in the context of this paper defines the post processing enhancement of medical images as opposed to 'high resolution' which defines native image resolution acquired during the MR acquisition phase. We also compare the SRGAN to three other models: SRCNN, SRResNet, and Sparse Representation. While the SRGAN results do not have the best Peak Signal to Noise Ratio (PSNR) or Structural Similarity (SSIM) metrics, they are the visually most similar to the original HR images, as portrayed by the Mean Opinion Score (MOS) results.