• Aktuelles

Vortragsreihe Medical Information Sciences (2 CME-Punkte)

Vortragsreihe MIS im Sommersemester 2023 im Hörsaal N2045 (FAI) Universität Augsburg (2 CME-Punkte)

Auf einen Blick:

Dienstag, 02.05.2023, 17:30 bis 19:30
Ort: N2045 (FAI)

Referent: Dr. Thomas Wendler Vidal (Technical University of Munich)

Titel: Using Anatomical and Physiological Constraints in Machine Learning for Medical Image Processing

Ort: N2045 (FAI) Universität Augsburg

www.uni-augsburg.de/de/fakultaet/fai/informatik/prof/bioinf/vortragsreihe-mis/

Der Vortrag richtet sich an ein interessiertes Publikum und werden in englischer Sprache gehalten. Im Anschluss an den Vortrag werden Snacks und Getränke serviert. 
 
Im Vorlauf der Vorträge wird zudem die Möglichkeit zur Wahrnehmung einer persönlichen Sprechstunde mit der oder dem Vortragenden des Tages angeboten, um sich bspw. über wissenschaftliche Fragen, Themen oder Kooperationsmöglichkeiten auszutauschen.

Bei Interesse bitten wir Sie, sich hierfür rechtzeitig über eine Nachricht an office.bioinf@informatik.uni-augsburg.de anzumelden.

 

Guest lecture series „Medical Information Sciences

We are delighted to announce that the guest lecture series Medical Information Sciences will continue this semester, once again covering a variety of topics relating to medical information sciences. It will be held on a weekly basis on Tuesday evenings at 5:30 pm during this summer term with alternating venues and live-streaming to the other location (University Campus and University Hospital Augsburg, respectively).

For the lecture on next Tuesday, May 2nd 2023, we cordially welcome

Dr. Thomas Wendler Vidal

(Director of the Interdisciplinary Research Lab at the Chair of Computer-Aided Medical Procedures and Augmented Reality, TUM, Munich)

 „Using Anatomical and Physiological Constraints in Machine Learning for Medical Image Processing“

Since the early 2010s, (deep) artificial neural networks started to flood the medical publication landscape, in particular in the fields of radiology, nuclear medicine and radiation oncology, by showing promising results in various tasks such as disease detection, outcome prediction, and therapy planning. Despite excellent reported performances, the deployment of such tools in clinical routine has been slower than expected due to the large variability in medical images.

Physics-informed machine learning has been proposed to increase robustness and improve the generalization of deep learning models. An extension of this concept in medical informatics is anatomy- and physiology-informed machine learning.

In the frame of the "Medical Information Sciences" lecture series, we present here different projects hosted at the Chair for Computer-Aided Medical Procedures at Technical University of Munich, where we use high-level knowledge of anatomy and physiology to constrain and regularize machine learning models, and as such, produce more robust medical image analysis tools.

 —————————————————————————————————————

Dr. Thomas Wendler Vidal is an Electronic Engineer (Technical University Federico Santa María, Valparaíso, 2004) with a Master of Science in Medical Technolgy (Technical University of Munich, 2007) and a Ph.D. in Computer Science (Technical University of Munich, 2010). After 9 years as CTO and CEO of medical device companies (SurgicEye GmbH, OncoBeta GmbH, ScintHealth GmbH), Dr. Wendler rejoined the Chair for Computer Aided Medical Procedures at Technical University of Munich as group leader at the Interdisciplinary Research Lab at Klinikum rechts der Isar. Dr. Wendler coordinates there the activities of the chair in the fields of Medical Image Processing and Robotic Ultrasound, as well as partnerships with the university hospital in the fields of Machine Learning and Computer-Aided Surgery. Dr. Wendler has authored >40 peer reviewed, has written two book chapters and has been granted 10 patent families (EU, US, DE). He is also currently the Vice-Chair of the Working Group for Digitalization and Artificial Intelligence at the German Society of Nuclear Medicine, and Comunication Officer of the Translational Molecular Imaging and Therapy Committee at the European Association of Nuclear Medicine.

Date: May 2nd 2023 (starting 5:30 pm)

Place: Hörsaal N2045 (Faculty of Applied Informatics)

 

Please note: The lecture will as an exception not be livestreamed to the IDM on next Tuesday.

We once again offer the chance to arrange informal individual meetings with the speaker to discuss related scientific questions and topics as well as common grounds with regard to personal research interests. Please send an e-mail to office.bioinf@informatik.uni-augsburg.de to make an individual appointment a few days in advance.

The lecture will last 45 to 60 minutes and will be followed by an open discussion. Each session will be rounded off with an informal get-together, where snacks and beverages will be served.

If you have any questions or notes, please feel free to contact us: office.bioinf@informatik.uni-augsburg.de