Artificial Intelligence (AI) for data-driven personalised medicine

This research project will develop explainable and interactive artificial intelligence (AI) for the analysis of large and complex clinical data, in order to improve respiratory medicine. On the one hand, we will explore and develop novel AI-based models for diagnostic and prognostic decision support at multidisciplinary board discussions, which are the gold standard for complex clinical decision making across medical disciplines. We will focus on AI tools that can assist multidisciplinary board discussions for interstitial lung diseases, which have a broad differential diagnosis. Particular emphasis of the AI solutions will be on multi-modal data integration and allowing superior performance through interaction between AI and the expert panels. On the other hand, we will develop AI tools for therapeutic decision support. We will rely on large clinical datasets to computationally extract phenotypes that can quantify therapeutic benefits of specific interventions. Mathematical models to predict individual treatment response from randomized controlled trials in COPD will be developed and clinical translation can be guaranteed by explaining model decisions and broad validation across datasets. As such, this project will be at the foreground of modern, data-driven personalized medicine.