Artificial intelligence (AI) can make a valuable contribution to the detection and treatment of intestinal polyps, but further research is needed to confirm findings and explore further possibilities. This is the conclusion of Elsa Soons in her dissertation, which she will defend on April 18 at Radboud University.
Soons and her colleagues found, among other things, that the use of AI to detect polyps led to a relatively high number of endoscopies in which at least 1 polyp was found. However, it turned out to be difficult for endoscopists to visually distinguish between neoplastic and non-neoplastic polyps during an endoscopy. AI may also be able to help with this in the future.
The second part of the thesis focuses on patients with familial adenomatous polyposis (FAP). Due to their greatly increased risk of colon cancer, a correct assessment of microscopic tissue is crucial for them in particular. Soons found that centers that more frequently assess tissue from FAP patients diagnose high-grade dysplasia almost 4 times more often than centers where this tissue is hardly assessed.
Soons conducted her PhD research at the MDL department of Radboudumc under the supervision of (co)promoters Prof. Dr. PD Siersema, Dr. MCA van Kouwen, Dr. TM Bisseling and Dr. RS van der Post.
Bron: