PARKINsensor

There's increasing demand for neurologists due to a growing number of patients and a severe shortage of specialists. Adequate alleviation of symptoms related to chronic movement disorders necessitates multiple follow-ups per year throughout a patient's lifetime, placing significant capacity demands on the healthcare system. Each follow-up is a complex task for the neurologist, who must discern longitudinal information including visual evaluation of movement, therapeutic doses, side effects and other related symptoms.

The Inspiration Behind the Innovation

The current gold standard for evaluating motor severity in Parkinson’s disease (MDS-UPDRS part III) is based on a visual assessment of motor tests examining rigidity, slowness of movement, tremor and balance impairment. The clinical scale has many drawbacks – for example, it’s time-consuming, limited by high variability and often fails to capture mild symptoms.

Motion capture using wireless IMU sensors has been used for decades in the entertainment industry for animation movies, and it’s a powerful tool for capturing movement. Adapting this tool for neurological assessment of motor severity makes it possible to accurately and objectively evaluate symptom severity at a refined level.  

The Innovation 

The team has designed a digital platform that visualises essential clinical data longitudinally in parallel with objective data about motor severity obtained from sensor-based motion capture. Although this solution can be adapted to several movement disorders, this project demonstrates its application for Parkinson’s disease – the fastest-growing neurological disorder worldwide and the second-most expensive to treat.

The innovation consists of algorithms that automate data analysis from sensor recordings, producing outcome measures that strictly align with clinical criteria. This solution could create a paradigm shift because its outcomes reflect true anatomical measures and are clinically relevant, enabling clinicians to tailor medication doses on a data-driven basis and with greater precision. Additionally, it can solve capacity challenges by enabling technical staff to perform routine motor tests, freeing up time for neurologists to focus on other crucial tasks.

Additionally, the solution produces high-quality, objective clinical data, providing a solid foundation from which to train AI algorithms in the future. Ultimately, AI could be used to recognise clinical patterns and provide decision-support features for clinicians, stratifying patients into subgroups to help neurologists select even more precise medication and doses.   

The Team

Jenny Ann Phan: MD, Post Doc; Regionshospitalet Gødstrup

Astrid Terkelsen: Consultant Neurologist, Clinical Professor; Department of Clinical Medicine – Neurology, Aarhus University Hospital; Aarhus University

Morten Stilund: Consultant Neurologist; Regionshospitalet Gødstrup

Erik Hvid Danielsen: Consultant Neurologist, Postgraduate Clinical Lecturer; Department of Clinical Medicine – Neurology, Aarhus University Hospital; Aarhus University