Ophthalmology

Diseases of the retina such as diabetic retinopathy or age-related macular degeneration are the most common causes of blindness in Germany. In recent years, there have been enormous developments in both the diagnosis and therapy of retinal diseases, which today enables us to offer many patients a tailor-made therapy. In the projects presented here, we are trying to address this variance of findings with the help of artificial intelligence and to develop more individualised diagnostic and prognostic tools for practitioners and patients.

Involved Research Groups

Prof. Dr. Andreas Stahl

Klinik und Poliklinik für Augenheilkunde
Universitätsmedizin Greifswald

Projects

  • Therapy prediction through analysis of patient data in ophthalmology: In this BMBF-funded project, we are investigating the course of treatment of patients with exudative age-related macular degeneration together with partners from informatics. The primary goal is to use AI-supported evaluation of image data and unstructured data from the patient’s medical record to process the large range of characteristics from individual factors and disease-associated image data in an information-technological manner. Intuitive visualisation in the form of a comparative explorative system enables the doctor to make a reliable prediction about the individual expected course of therapy for a patient. This should allow the therapy to be adapted more precisely to each individual patient. The ophthalmologist will be able to create a personalised treatment plan for each patient and minimise the risk of recurrence. For details see: https://topos.averbis.de/
  • Automated diagnosis of diabetic retinopathy:The gold standard in the diagnosis of diabetic retinopathy is funduscopy by a consultant ophthalmologist. Since the high number of patients with diabetes mellitus places high demands on availability and human resources, it would be desirable if the screening process could be made more efficient by using suitable tools. One starting point for this is the use of artificial intelligence on digital fundus photos of patients, which would allow patients to be presorted into categories such as “no diabetic retinopathy, mild, moderate or severe retinopathy”. This would (if appropriately proven reliable) greatly improve the processes of diabetes screening in the field of ophthalmology. In collaboration with the Department of Diabetology at the Karlsburg Clinic (Prof. Kerner), we are investigating the use of AI-assisted retinal diagnostics in patients with diabetes mellitus. The aim of this study is to compare funduscopy (standard diagnostics) with fundus photography and subsequent automated data analysis using artificial intelligence.

Publications & Conference papers

Stahl A.
The Diagnosis and Treatment of Age-Related Macular Degeneration.
Dtsch Ärztebl Int. 2020 Jul 20;117(29-30):513-520.

Grundel B, Bernardeau M-A, Langner H, Schmidt C, Böhringer D, Ritter M, Rosenthal P, Grandjean A, Schulz S, Daumke P, Stahl A
[Extraction of features from clinical routine data using text mining]
Der Ophthalmologe 2020 Jul 28. doi: 10.1007/s00347-020-01177-4

Berens P, Waldstein SM, Ayhan MS, Kümmerle L, Agostini H, Stahl A, Ziemssen F.
[Potential of methods of artificial intelligence for quality assurance].
Der Ophthalmologe. 2020 Feb 24. doi: 10.1007/s00347-020-01063-z. Review.

Bucher F, Mussinghoff P, Kühn T, Stahl A, Böhringer D.
[Technical aspects of quality assurance for intravitreal injections (IVI)].
Der Ophthalmologe. 2020 Jan 7. doi: 10.1007/s00347-019-01029-w. Review.

Lang GE, Stahl A, Voegeler J, Quiering C, Lorenz K, Spital G, Liakopoulos S. Efficacy and safety of ranibizumab with or without panretinal laser photocoagulation versus laser photocoagulation alone in proliferative diabetic retinopathy -the PRIDE study.
Acta Ophthalmol. 2019 Dec 6. doi: 10.1111/aos.14312.

Schmidt, C., Röhlig, M., Grundel, B., Daumke, P., Ritter, M., Stahl, A.,… & Schumann, H.
Combining Visual Cleansing and Exploration for Clinical Data
IEEE Workshop on Visual Analytics in Healthcare (VAHC) (pp. 25-32). 2019 IEEE.

Stahl A, TOPOs group
Using OCT to predict anti-VEGF treatment response and outcome.
DOG 2018, Workshop WS11: The future of retinal imaging –a peek behind the curtains

Holger Langner, Marc Ritter, TOPOs group
Applying Deep Learning methods to OCT image analysis in the context of Age-related macula degeneration.
DOG 2018, Symposium DS05: Deep learning in Ophthalmology –Technical approaches

Bastian Grundel, TOPOs group
TOPOS: Therapievorhersage durch Analyse von Patientendaten in der Ophthalmologie
DOG 2018, Workshop DS10: Deep Learning in der Augenheilkunde: Klinischer Einsatz