PhD, Université Paris Cité
Diagnosis support of rare genetic diseases: design of diagnosis support algorithms based on hybrid methods combining symbolic artificial intelligence and machine learning.
Rare diseases affect approximately 400 million people worldwide. Many of them suffer from delayed diagnosis. In this context, the objective is to integrate expert knowledge about the disease within machine learning models reusing patient data from electronic health records (EHR) to design a diagnosis support system. The model, that is meant to be used in a clinical context, must be reliable, explicable, and interoperable with EHRs.