At MEDICA 2019, learn firsthand about AI applications in the medical field from Dr. Esser, find out when they are coming, and what that really means at the MEDICA ECON FORUM by TK in November. In the run-up to the event, MEDICA-tradefair.com sat down with him for an interview.
Dr. Esser, your short presentation at the MEDICA ECON FORUM by TK examines whether AI will assist or rival physicians. Do you use AI in any way in your own work or research at this point?
Dr. Heinz-Wilhelm Esser: Yes, I do. In fact, I recently received an award for an application for the validation of tumor boards. Quality assurance in oncology has become very difficult as there are many available treatment options. Many choices lead to confusion, which is why patients are treated by experts in tumor boards after their cancer diagnosis. A tumor board is a group of doctors and other health care providers with different specialties that meet to discuss cancer cases and determine the best possible cancer treatment. Our application essentially enables us to make the decision of the tumor board experts and subsequently the tumor boards comparable.
This is a very exciting prospect, which is why I do not see AI as a rival or competition, but rather as an enhancement and complement.
How has it changed the way you work?
Esser: It is still too soon to tell. AI has not really made its way into the medical field yet. IBM Watson has tried and failed to deliver because its treatment recommendations were non-applicable, or Watson's decisions were suboptimal. This might have been the result of databases or programming. At any rate, I believe this still requires further extensive research. That being said, I think we can expect to see the first meaningful AI applications in the next five years.
What role should physicians play when it comes to the implementation of AI applications?
Esser: Physicians should still be the ultimate experts, populate the AI applications, set the action and review the findings. Needless to say, AI can essentially make a faster diagnosis than a human radiologist as it can access billions of images and compare the image at hand to make a medical diagnosis. Having said that, you still need quality assurance. I would personally not fully trust and rely on AI at this juncture.
Do you mostly see pros or cons of a potential use of AI in the medical profession?
Esser: I am certain that its use will ultimately give us an absolute advantage as doctors but we are still many years away from this scenario.