AI eye-image filter preserves disease signs while protecting patient identity
A medical informatics team at Jena University Hospital has introduced a privacy-compliant method for the AI-driven generation of realistic eye images. While the synthetic images remain suitable for diagnostic analysis, it is no longer possible to identify the individuals depicted in the original training data.
A team of medical informatics specialists at Jena University Hospital has developed a privacy-preserving approach to generate realistic eye images using artificial intelligence. The synthetic images retain diagnostic utility while removing individual identification markers. Presented in the journal PLOS Digital Health, the study addresses the challenge of balancing medical utility and anonymity in ophthalmology.
The researchers adapted a generative AI model to create synthetic eye images that closely resemble real ones, while altering biometric details to ensure anonymity. An analytical metric called the "Cone of Privacy" was used to filter out synthetic images with a higher re-identification risk. The method was tested on a dataset of 2,000 eye images from 704 individuals.
The algorithm successfully generated realistic synthetic images suitable for medical research and development, while effectively identifying and eliminating images that raised privacy concerns. The technique can also be applied to other medical image types, enabling secure data sharing and collaboration while adhering to data protection regulations.
Written by urgent.news from Medical Xpress's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
