The OPTIMISM project is managed by Steno Diabetes Center Copenhagen and SDU Robotics assists with tasks related to computer vision and image recognition. The project aims to make better use of current diabetic retinopathy screening resources in Denmark by, among other things, automating the task of analyzing retinal images from patients with diabetes.
Diabetic retinopathy is one of the most common complications in diabetes, and is considered the leading cause of blindness among working-age adults. An estimated 280,000 persons have diabetes in Denmark and yearly retinopathy screening is recommended for all diabetes patients. Screening is mainly conducted from retinal fundus photos with manual interpretation by specialized ophthalmological staff. The current screening methods are costly and time consuming, and as the demand for screening is expected to rise in the coming decades, patient wait times are likely to follow, which may lead to more cases of diabetes related blindness.
Most people with diabetes in fact show no signs of diabetic retinopathy at screenings, meaning that a lot of time is spent by healthcare professionals interpreting images without disease. By making use of deep artificial neural networks for automatic analysis of retinal images, the strain on medical staff, along wait times for patients, can be reduced, leading to better outcomes for those in more acute need for medical attention.
Funder: Novo Nordisk Foundation (NNF)
Project Partners:
- SDU Robotics, University of Southern Denmark, Denmark
- Steno Diabetes Center Copenhagen, Denmark
