Diabetic retinopathy screening in the emerging era of artificial intelligence – published online 31/05/2022

Jakob Grauslund

Using screening for the early detection of sight-threatening diabetic retinopathy is a pivotal step towards the reduction of visual loss in diabetes. In this issue, Jakob Grauslund (https://doi.org/10.1007/s00125-022-05727-0) presents the current state-of-the-art in diabetic retinopathy screening and outlines the start of the journey towards the adoption of new technologies and initiatives. These include handheld mobile devices, ocular telehealth programmes and automated image analysis using artificial intelligence. The author focuses on the clinical rationale and scientific evidence for deep learning, which has become the state-of-the-art in computer-based classification and segmentation in medical imaging. The author discusses how first regulatory approvals have been obtained for algorithms trained to detect sight-threatening diabetic retinopathy. He concludes that full-scale implementation in local and national screening programmes can be expected in the upcoming years, once ongoing challenges have been addressed in the transition from in silico experiments to clinical care. The figures from this review are available as a downloadable slideset.

All News