CD Laboratory for Artificial Intelligence in Retina

This CD Laboratory is researching AI-based systems for the diagnosis and monitoring of retinal diseases. The demographic-related increase in the number of cases and the demand for diagnostic precision make the use of AI systems unavoidable in the future.

Optical coherence tomography (OCT) is established as the diagnostic gold standard in the care of patients with retinal diseases. It enables the non-invasive and rapid acquisition of three-dimensional, high-resolution cross-sectional images and has a significant clinical and socio-economic value, with around 30 million examinations carried out worldwide every year. The unprecedented accuracy with which retinal changes can be phenotyped with OCT is offset by the challenge of managing and interpreting enormous amounts of image data for experts. In everyday clinical practice, the diagnosis of OCT examinations is generally qualitative and subject to the subjective judgement of the experts. For the holistic recording of an individual problem, it is expedient to merge the image data with the rest of the hospital file, but this is hardly possible on a comprehensive basis in everyday practice due to different documentation systems. Making clinical workflows more efficient in this respect in future is of great relevance, as the increasingly older population and rising prevalence of diabetes, for example, are expected to pose massive challenges in relation to the care of patients with age-related macular degeneration (AMD) and diabetic retinopathy. Forecasts predicting 288 million patients affected by AMD in 2040 emphasise the urgency of the problem.

Since the advent of deep learning, artificial intelligence-based systems in the field of Medicine diagnostics and prognostics have come close to the capabilities of human examiners and even surpass them in certain areas. They enable the extraction of quantifiable biomarkers from Medicine image data and thus have the potential to raise both the diagnosis and monitoring of retinal diseases to an unprecedented level of precision. In addition, the possibility of discovering new subtypes of diseases through large-scale, data-based approaches should also be mentioned. The potential of image data analysis using artificial intelligence must therefore be put into practice.

The aim of this CD Laboratory is to develop a clinical support tool based on artificial intelligence to equip a standard OCT device that supports clinicians in the early detection, classification of subtle pathological changes and management of individual disease progression. The quantity and variety of data required for this purpose will shift the focus to methodological approaches based on widely available "real-world" data. This CD Laboratory will bring together interdisciplinary experts in the field of Medicine image data analysis and retinology and pioneers in the application of artificial intelligence in OCT analysis.