KINDA (KINematic DBS Algorithm)

This project is a collaboration between Amsterdam UMC, Medtronic, and Machine Medicine, forming a strong public-private partnership aimed at automating the most complex part of Parkinson’s treatment: the adjustment of deep brain stimulation (DBS). By combining new DBS electrodes that can record brain activity with artificial intelligence (AI)-based video analysis of symptoms, we aim to develop a smart, adaptive form of DBS that responds automatically to a patient’s individual needs. 

Parkinson’s disease (PD) is the world’s fastest-growing neurological disorder and affects over 50,000 people in the Netherlands alone. It is projected that this number will double by 2040. In the advanced stages of the disease, DBS is the first-choice treatment according to international guidelines. While it significantly improves quality of life, DBS programming is currently a laborious, specialist-driven process that o[en requires multiple hospital visits. As the number of patients grows, waitlists increase and healthcare resources become strained. There is a pressing need for more efficient, personalized care. 

In this study, 50 patients with Parkinson’s disease will receive DBS electrodes that can both stimulate and record brain signals. In the weeks before and months after surgery, brain activity and symptoms will be recorded using wearable sensors, clinical assessments, and AI-based video analysis. These data will be used to develop AI models that identify links between brain activity and individual symptoms. In the final phase, the models will be used to automatically adjust the DBS settings in real time. 

Deliverables of the project include AI models that predict Parkinson’s symptoms based on brain activity, validated neural biomarkers for symptom control, and clinical data on the safety and effectiveness of adaptive DBS. This innovation could make DBS more effective, reduce the need for specialist intervention, and improve access to care for the growing number of patients.