Prof. Maarten de Vos from KU Leuven shares insights into how AI and digital biomarkers are reshaping Parkinson’s disease care, a key focus of the AI-PROGNOSIS project. Traditional diagnosis and monitoring methods rely on periodic clinical assessments, which may not fully capture symptom fluctuations or individual differences. AI-PROGNOSIS aims to bridge this gap by leveraging everyday digital technologies to improve detection, assessment, and prognosis.

The challenge of traditional care

Parkinson’s disease is often diagnosed based on visible symptoms, leading to delays in detection. Once diagnosed, patients typically visit clinicians every six or twelve months for checkups. However, these assessments only provide a snapshot of a person’s condition at a single point in time. Given that symptoms can fluctuate and vary significantly from person to person, this approach has limitations.

The role of AI and digital biomarkers

AI-PROGNOSIS explores how digital biomarkers—measurable indicators of disease progression derived from digital technology—can enhance Parkinson’s care. By using widely accessible tools like smartphones and smartwatches, researchers can collect real-world data on movement patterns, voice changes, and other key markers. AI-driven models then analyze this data to:

• Improve early diagnosis;

• Assess disease severity more objectively;

• Predict disease progression over time.

From research to clinical implementation

Developing AI models is just one part of the solution. Ensuring their real-world application is equally critical. AI-PROGNOSIS prioritizes rigorous validation of these models to demonstrate their reliability and clinical value. Collaboration with patients and healthcare professionals is essential to integrating these technologies into routine practice, making them both accessible and effective.

By advancing AI-powered monitoring and prediction, AI-PROGNOSIS is working towards a future where Parkinson’s disease care is more proactive, personalised, and data-driven.