When GANs meet radiology: a curated paper trail

What it does This repository is a curated bibliography—not a codebase—collecting academic papers that apply generative adversarial networks to medical imaging. It organizes research into clinical tasks such as low-dose CT denoising, segmentation, lesion detection, image synthesis, reconstruction, and registration. Each entry links out to Google Scholar, arXiv, or publisher pages, with occasional pointers to implementation code.

The interesting bit Rather than treating GANs as a generative novelty, the list tracks their migration into clinical pipelines: reducing radiation exposure by denoising scans, synthesizing data to bypass annotation bottlenecks, and segmenting organs across MRI, CT, and fundoscopic images. It is essentially a literature review in Markdown form.

Key highlights

• Covers eight distinct categories, from review articles to image registration
• Heavy focus on CT denoising and MRI segmentation, with entries spanning brain tumors, retinal vessels, and cardiac structures
• Mixes peer-reviewed journals (TMI, MedIA) with conference papers and arXiv preprints
• A minority of papers include direct links to companion code repositories
• Maintained in the “awesome-list” tradition for discoverability

Caveats

• This is strictly a reading list; the repository itself contains no executable code or models
• The provided README truncates partway through the Detection section, so the full scope of the list is not visible
• Maintenance status and update frequency are unclear from the sources

Verdict Worth bookmarking if you are a researcher or clinician navigating where adversarial methods have been tried in medical imaging. Skip it if you need a unified library or ready-to-run implementation.