Eye2Gene AI Technology Streamlines Inherited Retinal Disease Diagnosis

UK researchers from the UCL Institute of Ophthalmology and Moorfields Eye Hospital have introduced Eye2Gene – an AI technology that predicts which genes may be impacted by pathogenic variants in patients with hereditary retinal diseases. Remarkably, it does not need to rely on human phenotype ontology (HPO) terms. The purpose of Eye2Gene is to simplify diagnoses of this lifelong condition, which remains undiagnosed for almost 40% of patients in the UK, presenting a significant health and economic burden. A genetic diagnosis can also impact matters such as prognosis and family planning choices. Eye2Gene was developed using data from over 1,000 patients before being tested on 130 patients with known genetic diagnoses and retinal data. In up to 75% of cases, the predicted gene accuracy was the same as or better than that produced by the HPO-based Exomiser programme. More evaluations and clinical trials will take place before Eye2Gene is deployed within healthcare clinics as a piece of medical device software.