Authors: John R. McNulty, Lee Kho, Alexandria L. Case, Charlie Fornaca, Drew Johnston, David Slater, Joshua M. Abzug, Sybil A. Russell

Published on: March 28, 2024

Impact Score: 8.4

Arxiv code: Arxiv:2403.19107

Summary

• What is new: Development of the GAN Image Synthesis Tool (GIST), an open-source synthetic image generation pipeline focusing on medical imaging for conditions with low incidence rates.
• Why this is important: Limited access to medical imaging data due to privacy concerns and the rarity of some diseases.
• What the research proposes: A synthetic image generation pipeline, GIST, utilizing Generative Adversarial Networks (GANs) to produce high quality, clinically relevant synthetic images.
• Results: The pipeline successfully generates high quality, clinically relevant knee and elbow x-ray images evaluated through layperson’s assessment and the Fréchet Inception Distance metric.

Technical Details

Technological frameworks used: GAN Image Synthesis Tool (GIST), based on Generative Adversarial Networks

Models used: Evaluates current GAN architectures

Data used: Radiography, specifically knee and elbow x-ray images

Potential Impact

Digital health space, AI diagnostic tools, healthcare providers, medical imaging software companies

Want to implement this idea in a business?

We have generated a startup concept here: SynthMed Imaging.