Detection of adhesions on cine-MRI with artificial intelligence

Wednesday 22 May 2024, 10:30 am
PhD student
B. de Wilde ir.
Promotor(s)
prof. dr. ir. H.J. Huisman, prof. dr. H. van Goor
Co-promotor(s)
dr. R.P.G. ten Broek
Location
Aula

Annually, 30,000 patients experience chronic abdominal pain after surgeries, often caused by internal scar tissue known as adhesions. Cine-MRI is a novel technique capable of detecting adhesions, thereby preventing unnecessary surgeries. However, assessing cine-MRI is time-consuming and requires expertise. Artificial intelligence can learn adhesion detection, reaching a level comparable to a radiologist without specific experience in cine-MRI. Additionally, AI aids inexperienced radiologists in assessing cine-MRI, approaching expert radiologists' proficiency. Consequently, this technique can be more readily implemented in multiple hospitals, ultimately leading to enhanced care for patients with chronic abdominal pain.

Bram de Wilde (1996) received his double Master's degree in Applied Physics and Applied Mathematics from the University of Twente, Enschede, in 2019. In October 2019 he started his PhD in the Diagnostic Image Analysis Group of the Radboud university medical center, Nijmegen, working on AI-assisted detection of adhesions on cine-MRI. Currently, he is working as a postdoctoral researcher at the University of Twente.