Exploring Parkinson's Disease Datasets: Key Findings, Challenges, and Recommendations for Motor Symptom Analysis.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Data Science

Background:

• Parkinson's Disease (PD) is a neurodegenerative disorder with characteristic motor symptoms like bradykinesia and tremor.
• Advancements in sensing technology have enabled data collection for PD motor symptom analysis.
• Machine Learning (ML) and Deep Learning (DL) show promise for early diagnosis and personalized treatment in PD.

Purpose of the Study:

• To survey widely used Parkinson's Disease motor symptom analysis datasets.
• To examine dataset features, modalities, and data sources.
• To address challenges related to dataset variability and accessibility.

Main Methods:

• Systematic review of 17 prominent Parkinson's Disease motor symptom datasets.
• Analysis of dataset characteristics including features, data collection methods, and patient populations.
• Identification of common challenges and limitations across datasets.

Main Results:

• Compilation of key features and modalities from 17 PD motor symptom datasets.
• Identification of significant variability in data collection and features across datasets.
• Highlighting challenges in patient accessibility and dataset availability for research.

Conclusions:

• Standardization of datasets is crucial for advancing ML/DL applications in Parkinson's Disease.
• Addressing data accessibility and variability will facilitate cross-dataset studies and wider implementation.
• Further research is needed to overcome current limitations in PD dataset utilization.