Past Event! Note: this event has already taken place.
Seminar: AI for ultrasound-guided medical procedures and competency-based education
March 9, 2020 at 1:30 PM to 2:30 PM
|Location:||5345 Herzberg Laboratories|
Data Science Distinguished Speaker Seminar Series
Refreshments will be provided
Deep learning has changed the way we process images and videos today. Artificial neural networks enabled more accurate image analysis methods than we could ever design before. This new family of methods is yet to reach its full potential in clinical applications. The deep learning research community has shared their software tools for neural networks as free, open-source tools. To facilitate the adoption of these methods in medical procedures, the biomedical computing community also need to work towards a robust open-source platform for deep learning in medical applications.
The collection and curation of large-scale, high-quality training data is the key to success in AI-based projects. But it requires more collaboration from our research community than any earlier technology. The Laboratory for Percutaneous Surgery (Perk Lab) has been developing and maintaining open-source software tools for minimally invasive medical interventions for over a decade. Our software platform has become the most used research platform in navigated medical interventions research. Building on this tradition, we aim to facilitate AI-guided medical interventions by providing shared tools for data collection, data management, annotations, and data sharing for the research community. Proof of concept examples will be presented on applications of AI-based ultrasound image analysis and automatic feedback in surgical skills training by real-time video processing.
About the Speaker
Tamas Ungi received his MD in 2006 and PhD (Radiology) in 2011 from the University of Szeged, Hungary. He is currently an Adjunct Assistant Professor and Senior Research Manager at Queen’s University in Kingston, Ontario with appointments at the School of Computing and the Department of Surgery. He is responsible for the translation of new technologies to clinical research and commercial applications. His research interests are image-guided medical interventions, interventional skills education, and real-time applications of artificial intelligence in these procedures. He has been leading the development of open-source software tools for medical interventions and for procedural skills education. Dr. Ungi has authored over 50 journal papers, 150 conference papers, 4 patents, and two book chapters. He has organized several tutorials on how to use open-source tools for accelerating research.