DIGIPD : Validating DIGItal biomarkers for better personalized treatment of Parkinson’s Disease

The aim of the project is to evaluate in how far Digital biomarkers (DMs) extracted from a mobile gait sensor system as well as recordings of voice (via phone) and face movement (via video) could help reaching accurate disease diagnosis and treatment dependent prognosis for each individual patient with Parkinson’s Disease (PD). Starting from pre-existing data and advanced AI methods developed by project partners in the past, it assesses different types of DMs regarding their ability to discriminate between PD and healthy controls, for predicting different types of PD disease trajectories and for predicting the treatment dependent change of disease symptoms over time. It also investigates via statistical and AI methods the relation of different types of DMs to each other, to clinical outcome scores and to molecular disease mechanisms, hence opening the opportunity to a new level of interpretation of DMs. Finally, it provides important insights into feasible legal pathways for the use of AI and DMs in the field (specifically with respect to data privacy aspects), and investigates the acceptance, ethical concerns and perceived value of such approaches by patients.

Researchers : Noémi Bontridder