Skip to main content English



26. Juni 2017
12:00 - 13:30

16 G (roter Bettenturm)
AKH Wien
Währinger Gürtel 18-20, 1090 Wien


12:00-12:15:        Univ.-Prof. Dr. med. Ursula Schmidt-Erfurth: Einleitung und Vorstellung

12:15-12:45:        Priv.-Doz. Dr. med. univ. Sebastian Waldstein, PhD

13:00-13:30:        Dr. Bianca Gerendas, MSc



Machine learning – neue Konzepte aus der ophthalmologischen Bildanalyse

Priv.-Doz. Dr. med. univ. Sebastian Waldstein, PhD
Univ.-Klinik für Augenheilkunde und Optometrie

DF-Punkt: 1

The introduction of high-resolution, high-speed optical coherence tomography (OCT) has revolutionized diagnosis and management of retinal diseases, the leading cause of blindness in developed countries. At the same time, modern OCT generates prohibitively large amounts of imaging data. The sheer scale of these data limits evaluation by human observers in both clinical and scientific practice, but offers opportunities for artificial intelligence-based analysis and research. This includes personalized medicine approaches to predict individual disease risk and therapeutic response.

This contribution will provide an overview of current developments in the area of computational analysis of retinal imaging. It will introduce the two basic concepts of machine learning, i.e. supervised and unsupervised learning. Automated detection and quantification of biomarkers in retinal images based on supervised learning, as well as the development of predictive algorithms will be discussed.

The opportunities to make novel discoveries with massive OCT image datasets (“Big Data”) are beginning to be realized with unsupervised machine learning using anomaly detection. This unbiased analysis of OCT data is capable to identify the pathomorphological fingerprint of retinal disease and is therefore most promising to predict individual visual function and patient response patterns. This enables the discovery of novel imaging biomarkers that are highly predictive of visual function and morphologic as well as functional response to advanced retinal therapies. The concept of spatiotemporal disease modeling allows the generation of personalized prognostic tools to combat the leading sight-threatening diseases.

Sebastian Waldstein is an ophthalmologist and retina specialist at the Department of Ophthalmology at the Medical University of Vienna. He obtained an MD degree from Innsbruck Medical University in 2011 and a PhD in Medical Physics from the Medical University of Vienna in 2017. He was appointed “Privatdozent” in ophthalmology in 2016.

Sebastian Waldstein is Associate Director of the Christian Doppler Laboratory for Ophthalmic Image Analysis, an interdisciplinary research group that has established itself as one of the leading labs in ophthalmic image analysis worldwide. He has extensive experience in clinical ophthalmology and in the field of machine learning applied to medical imaging, having published >40 peer reviewed journal and conference publications and 2 book chapters in the subject field. He is PI of several industry-funded grants for machine learning in OCT of age-related macular degeneration.


 The Vienna Reading Center – Standardized and Independent Image Analysis

Dr. Bianca Gerendas, MSc

Univ.-Klinik für Augenheilkunde und Optometrie

DF-Punkt: 1

The Vienna Reading Center (VRC) provides imaging services for clinical, pharmaceutical and academic research, until now focussed on ophthalmology. We are a professional team of 25 people equipped to respond to the requirements of both international multi-center and single-site investigator studies with a focus on multimodal image analysis. Our group includes dedicated clinical experts, high-end custom software developments with year-long ophthalmic image analysis background, especially for “Big Data”, a large computational imaging cluster, administrative and quality management personnel and certified readers.

The VRC has established a solid network of collaborations with over 200 VRC certified academic sites distributed across 5 continents, which have provided images successfully through a live, online, automatically quality-controlled platform for numerous prospective randomized clinical trials, many of them phase IIIb clinical trials licensing new drugs in Europe. Our services are supported by state-of-the-art technology offering conventional as well as the most specialized methods available in image analysis. Founded in 2005 as the first fully digital image analysis platform for ophthalmology world-wide, the VRC is a subunit of the Department of Ophthalmology. An important focus of our group is the development of computational image analysis methods and its implementation in our systems. This activity is fostered by close collaboration with the Christian Doppler Laboratory for Ophthalmic Image Analysis (OPTIMA) and mirrored in our numerous publications in this field.

With this 12 year background in standardized and independent image analysis (including manual, semi-automated and fully-automated tasks) and a flexible platform, as well as software and databases with ISO certification, meeting all relevant FDA and EMA requirements (several successful audits/monitoring visits), we would now like to extend our services to other departments within the Medical University of Vienna with relevant tasks in medical image analysis. This talk will give an overview of the achievements of the VRC both in services as well as our research outputs and will be the first introduction of the “interdisciplinary” VRC.

Bianca S. Gerendas is the Director of the Vienna Reading Center and Associate Director of the Christian Doppler Laboratory for Ophthalmic Image Analysis (OPTIMA). She completed Medical School in 2009 at the University of Heidelberg, Germany, where she in parallel obtained her Master of Science in Healthcare Management. Bianca Gerendas completed her doctoral thesis before joining the Medical University of Vienna in 2010 and is currently finishing her PhD in Medical Physics in parallel to her residency training in Ophthalmology; she recently submitted her “Habilitation in Ophthalmology” and is currently awaiting its approval. As PI of several grants both from international peer-reviewed projects as well as industry funding, her primary research interests are in the field of Innovative and Interdisciplinary Computational Ophthalmic Image Analysis, Translational Ophthalmic Research, Retinal Structure-Function Correlation, Large Scale Data Analysis/Big Data and Imaging Biomarkers which can be seen by her >30 peer-reviewed journal publications and conference proceedings, 3 book chapters, >100 congress presentations and invited talks as well as numerous awards in her research field.