Improved Diagnosis at Radboudumc using Artificial Intelligence

Thomas van den Heuvel, postdoctoral researcher at Radboudumc, found a method to improve the procedure of diagnosing hip dysplasia in children using Artificial Intelligence (AI). Hip dysplasia is diagnosed using an ultrasound examination but pediatricians often lack the knowledge on how to perform such procedure. By combining ultrasound with AI support, Thomas was able to train pediatricians to make a diagnosis in one hour and therefore preventing a lot of unnecessary care.

At the pathology department, Jeroen van der Laak, professor in Computational Pathology, is also developing AI-based applications to improve healthcare. Jeroen expects that AI will play a major rol in diagnostic procedures (e.g., cancer diagnosis). Jeroen says that AI can be used for more consistent diagnoses and to predict if a treatment has a positive effect on a new tumor. This makes it possible to prevent unnecessary treatment and healthcare costs.

Jeroen mentions that AI is not perfect yet and that a human expert should still make the final diagnosis. AI can still make mistake and should therefore be used as a support tool rather than a replacement of the human expert. Just like with medicine, approval of AI applications is very strict. But both Jeroen and Thomas expect that AI will be applied on a large scale in the healthcare system in the future.