PhD defence by Frederik Jager Bruun
Understanding Clinical Limitations of AI-Based Fracture Detection on Radiographs
Assessment Committee:
Associate professor Jonathan Frederik Carlsen, Department of Clinical Medicine, University of Copenhagen (Chairperson)
Professor Jeppe Lange, Aarhus University
Dr Susan Shermeldine, University College London
Supervisors:
Clinical Professor Mikael Ploug Boesen
Postdoc Mathias Willadsen Brejnebøl
Clinical Associate Professor Philip Hansen, Postdoc Christoph Felix Müller, Forskningsradiograf Janus Uhd Nybing,
Department:
Department of Clinical Medicine
Graduate Programme:
Basic and Clinical Research in Musculoskeletal Sciences
Place:
Fælleshuset, Room: Lassen auditoriet, Nielsine Nielsens Vej 12, 2400 København
Email address to gain access to the thesis: jagerbruun@gmail.com
You will either receive a copy of the thesis or be informed where you can read a physical copy.
Recipients of copies of the thesis are not allowed to share or distribute it due to copyright compliance.
Short description of the thesis:
One of the most prominent applications of artificial intelligence is image analysis. Within healthcare, AI tools for diagnostic imaging have been developed and implemented in hospitals worldwide. Fracture detection on radiographs is among the most widely adopted applications and is now used in most Danish regions. But how well does the fracture detection algorithm implemented in the Capital Region of Denmark actually perform? And what systematic errors does it still struggle to overcome?