Skip to content

AI in medicine – knowledge is in demand

How can people remain at the center of the medicine of the future when a lot of data is brought together, chatbots act as coaches and digital twins support prevention and therapy?

Author: Claudia M. Witt

Artificial intelligence (AI) will become part of medical care – but how far should integration go? AI attempts to replicate understanding and learning with software, in that the systems (for example, as neural networks) learn skills from large amounts of data, which can then be used for predictions, for example. This is also how search engines on the Internet work.

AI is also increasingly being used in medicine. My health insurance company is currently offering me a digital insurance model with a symptom checker based on AI. My first reaction: How exciting – I want to learn more about that so I can make an informed decision. But it wasn’t that easy, because my health insurance company couldn’t answer any of my questions about the AI algorithm. For instance, I wanted to know: What is the algorithm used for? How was it trained? How was the “bias” reduced, that is, how good are the predictions for different people? What do I do if I think the AI result is wrong? That the topic of “AI in medicine” is in right now, and that education about AI can currently be a problem, is known by others. For example, this fall the Swiss Medical Association (FMH) published clear demands on the use of artificial intelligence.

The developments in digital medicine are rapid. Therefore, it is necessary to look into the future, so that we do not run behind the development, but can help to shape a reflected medicine of the future – in which people remain the focus.

Claudia M. Witt, MD, MBA, is co-director of the Digital Society Initiative (DSI) at the University of Zurich (UZH) and Professor of Complementary and Integrative Medicine at the UZH Faculty of Medicine. She leads diverse projects in the field of Digital Health.