Breakthrough in predicting cause of Inherited Retinal Diseases (IRDs) with help of AI.
A new artificial intelligence tool, Eye2Gene, developed by researchers at University College London (UCL) and Moorfields Eye Hospital, has been shown to accurately predict the genetic cause of inherited retinal diseases (IRDs) directly from routine eye scans. The study, published in Nature Machine Intelligence, marks a major step forward in accelerating genetic diagnosis for IRDs—one of the leading causes of blindness in working-age people. To this day, patients affected by these conditions face diagnostic odysseys.
Rare eye diseases such as inherited retinal diseases (IRDs) are challenging to diagnose genetically. IRDs are typically monogenic disorders and represent a leading cause of blindness in children and working-age adults worldwide. A growing number are now being targeted in clinical trials, with approved treatments increasingly available. However, access requires a genetic diagnosis to be established sufficiently early. Critically, the timely identification of a genetic cause remains challenging.
To combat this issue, Professor Nikolas Pontikos has developed a deep learning model, Eye2Gene, able to predict the causative IRD gene from the retinal scans of a patient with an IRD. Eye2Gene was trained on retinal scans acquired in individuals with IRDs seen at Moorfields Eye Hospital who had undergone genetic testing and where a confirmed genetic cause had been identified by an accredited diagnostic laboratory. Eye2Gene was internally validated on a held-out set of retinal scans from the eye hospital, and externally validated on retinal scans from individuals with IRDs from five different external clinical centres. Eye2Gene performance was also compared to that of expert clinicians.
The Eye2Gene predictions were shown to outperform human expert-level accuracy at a rate of 76% when tested, whereas ophthalmologist assessments achieved an average top-five accuracy of 29.5%.
This breakthrough was made possible thanks to vital funding from Sight Research UK, whose continued support enabled the research underpinning Eye2Gene’s development. The model was trained on one of the world’s largest IRD imaging datasets and validated internationally across five clinical centres.
Associate Professor Nikolas Pontikos and Professor Michel Michaelides who lead this study said:
“We are incredibly grateful to Sight Research UK for their investment in this work. Their support is instrumental in helping us move from an idea to a real-world tool that can assist clinicians and ultimately benefit patients.”
By prioritising likely gene diagnoses and helping to guide genetic testing, Eye2Gene could significantly reduce the diagnostic journey for people with inherited sight loss and improve access to gene-based treatments and trials.
