The project “SEDAAR” (Screening for Eye Disease using Automated Analysis of Retinal Photography) submitted by ADCIS and the Hospital of the Adolphe de Rothschild Foundation was chosen from 138 submissions as one of the10 winners of the “Artificial Intelligence for an improved healthcare system experience” challenge organized by the HDH organization (Health Data Hub) and BPI France. This was the only project submitted in the field of ophthalmology.
Fully integrated at the beginning of the vision care pathway, SEDAAR has three main goals:
- Proved access to mass screening for the most frequent retinal pathologies;
- Target patients who require rapid medical attention;
- Build a large database of color fundus and OCT images for public and/or private R&D.
During the course of this project, a well-organized database of over 95,000 fully annotated OCT and color fundus images will be generated. This multi modal and multi pathology database will be made available to the scientific community. It will allow AI researchers and specialists to develop the best approach to assist in decisions and to the interpretation of retinal images being performed by orthoptists and ophthalmologists. In addition to development work already completed for the detection of diabetic Retinopathy, ADCIS will develop a new set of algorithms based on Deep Learning technology.
ADCIS will grow and strengthen its expertise in AI applied to ophthalmology during the 3 years of this project.