Study Rationale:
The use of wearable sensors to detect and measure Parkinson’s disease shows great promise. However how to convert the raw data to specific measures is poorly understood. We propose a series of data challenges to develop new robust and validated measures from wearable sensors.
Hypothesis:
Measures of Parkinson’s symptoms and severity from digital wearable sensors can be improved, as measured by an unbiased metric, through collaborative challenges.
Study Design:
Data from past MJFF-sponsored trials using digital wearable sensors will be gathered together and evaluated for suitability to be included in a challenge. One or more specific questions around measurement of Parkinson’s disease will be chosen for challenge participants to address. The data will be split into a challenge and evaluation set. Researchers from around the world will then be invited to develop measures from the challenge set, and we will evaluate these measures in the evaluation set to choose the best measures.
Impact on Diagnosis/Treatment of Parkinson’s Disease:
Best measures can be used in future trials and possibly for at-home monitoring of disease.
Next Steps for Development:
This is the first step in developing and evaluating novel measures of disease severity and symptoms. A requirement of participation in the challenge is the sharing of how these measures are derived so that anyone can re-use them in their own studies.