Objective: To develop and evaluate deep-learning models for predicting Parkinson’s disease using brief-action motion-sensor data from sit-to-stand and turning tasks.To develop and evaluate deep-learning models for predicting Parkinson’s disease using brief-action motion-sensor data from sit-to-stand and turning tasks.

Background: Early initiation of lifestyle and pharmacologic treatments can enhance the quality of life, reduce costs, and improve outcomes in patients with Parkinson’s disease (PD). Currently, few biomarkers exist that can help with early diagnosis. Moreover, they are hindered by testing costs, invasiveness, and lead time. Motion sensor data effectively captures spatial movement patterns, which can serve as a biomarker collected in the clinic, community, or resource-limited and rural settings where medical practice is dictated primarily by generalists. We propose a deep-learning model using brief action motion-sensor data to predict PD.

Method: We utilized data from 403 sit-to-stand and 1749 turning episodes. We built neural-network models using sit-to-stand, 2D, and 3D turning data and an ensemble model using sit-to-stand and 3D turning data. SHAP (SHapley Additive exPlanations) values were calculated to identify key features influencing predictions.

Results: The sit-to-stand model demonstrated an accuracy (±standard error) of 77% ± 1.3%, precision of 74.7% ± 1.9%, sensitivity of 80.5% ± 1.8%, and specificity of 73.6% ± 1.9%. The 2D and 3D turning models and the ensemble STS and 3D turning models had accuracies of 69.2%, 72%, and 76.6%, respectively. SHAP analysis revealed that the whole episode duration was the most influential feature in the STS model, and the number of turning steps for both 2D and 3D turning models consistently explained the model’s predictions.

Conclusion: Our models using motion-sensor data perform well in detecting Parkinson’s disease during sit-to-stand and turning tasks. With augmented data to improve generalizability, these models have the potential for deployment in resource-limited clinical settings, enhancing personalized medicine and significantly improving quality of life through earlier referrals and treatment initiation.

Motion sensor positions

Receiver operating characteristic curves

Shapley additive explanations for the AI models

« Back to 2025 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/neural-network-models-predict-parkinsons-disease-using-brief-action-motion-sensor-data/