Neurological disorders are now the leading cause of disability worldwide. Parkinson's disease is the second most common neurodegenerative disorder, and case numbers are growing quickly. The disease presents a great challenge for people living with it, caregivers, relatives, and society as a whole. Despite decades of scientific effort, there is still no cure, and diagnoses are often made only in advanced stages of the disease. Emerging artificial intelligence (AI) methods and digital technologies offer new possibilities for improving early and accurate diagnosis and tracking of disease progression. These techniques could help support a better personalized treatment of the disease in the future. For instance, AI-based analysis of voice recordings – already a focus of active research – shows great potential for providing early indications of symptoms that are typical of Parkinson's disease.

Voice biomarkers are just one focus of the new Marie Skłodowska-Curie Doctoral Network, Artificial Intelligence in Parkinson's Disease (AIPD). 14 doctoral candidates will work on groundbreaking methods to improve early detection and treatment strategies for Parkinson's disease. Questions surrounding the trustworthiness, ethics, and regulation of AI solutions will also be a central focus of the research work.

The AIPD network brings together leading academic institutions, pharmaceutical companies, and medical technology enterprises across Europe. In addition to Fraunhofer SCAI, prominent academic partners include the universities of Bonn, Luxembourg, Namur, and Pisa, the Erasmus Medical Center in Rotterdam, the Luxembourg Institute of Health, and the RCSI University of Medicine and Health Sciences in Ireland. Key industrial partners such as Novo Nordisk, GE Healthcare, petanux, ki:elements and the Centre Hospitalier de Luxembourg play also a leading role. Extended research stays with industrial partners provide unique interdisciplinary training that bridges academic and industry perspectives.

Fraunhofer SCAI coordinates the project and contributes its extensive expertise in data-driven methods in biomedicine. This includes AI-based techniques for simulating individual digital twins of Parkinson's patients and approaches to improving predictive models of disease progression. In this way, explainable AI models are created that are tailored to the individual patient.

Project duration: 11/2024 to 10/2028

More info