Clinical Decision Support Systems: Beyond Model Performance
Translating advances in machine learning in clinical practice needs to take into consideration not only model performance but also model explainablity, fairness, internal and external validation, bias and uncertainty estimation as well as privacy of the patients. Furthermore, this information should be provided to clinicians and patients in a timely and intuitive way. In this talk, I am going to explain why a Human-Centered Artificial Intelligence (AI) framework is important and the challenges involved in developing it. I am going to briefly review recent guidelines in healthcare for AI and emphasize common misconceptions.
