Within the medical curriculum we are very active in the second track module dedicated to “digital health”. The program encompasses teaching of central concepts in machine learning and AI as well as applied courses on data integration and analyses.
For instance, we explain how the performance of AI algorithms is evaluated, with an (unfortunate) example of clinical relevance: COVID19.
Furthermore, we introduce the analysis of genomic data, in particular state-of-the-art approaches, such as data augmentation and imputation (“simulate what is not measured”). Deep learning-based image analysis is taught, too. And the integration of various data types, risk prediction, plus a critical discussion of robustness and explainability are essential skills to the next generation of AI experts in biomedicine.