EviRed, a new artificial intelligence algorithm, is being ‘trained’ to predict the risk of diabetic retinopathy (DR) using modern imaging techniques.
Professor Ramin Tadayoni at the University of Paris, France, is leading the Intelligent evaluation of diabetic retinopathy (EviRed) project, which is intended to improve upon the Airlie House system. Although the latter method of evaluation was enhanced by the Early treatment diabetic retinopathy study (ETDRS) research group, Prof Tayadoni said, “The ETDRS classification is basically an estimation of the risk of progression to proliferative DR based on findings from stereoscopic seven-field fundus photographs. It was a wonderful system when it was first developed, which was at a time when nothing else was available besides fundus photography, but it provides insufficient prediction precision for modern patient care.”
EviRed will factor ultra-widefield photography, OCT and OCT angiography, plus clinical factors such as blood pressure and glycaemic control, into its predictions. The use of OCT alone may also play a role in predicting the onset of diabetic macular oedema (DMO).
In France, a trial of 5,000 diabetic adults is underway to evaluate EviRed as a prognostic tool. Its accuracy in assessing DR severity extrapolating progression to severe retinopathy will be measured against the predictions of ophthalmologists using the ETDRS classification system.