Analyzing Gait Pattern Associated With Neuropsychiatric Symptoms in Parkinson's Disease by a Comprehensive Approach

Area of Science:

• Neurology
• Biomedical Engineering
• Clinical Diagnostics

Background:

• Parkinson's disease (PD) presents with motor and non-motor symptoms, including neuropsychiatric symptoms (NPS) like depression and anxiety.
• NPS significantly impact the quality of life for PD patients.
• Understanding the link between motor and non-motor symptoms is crucial for comprehensive PD management.

Purpose of the Study:

• To investigate gait patterns in PD patients with and without NPS.
• To analyze kinematic, kinetic, and spatio-temporal gait variables.
• To assess the potential of gait analysis as a biomarker for NPS in PD.

Main Methods:

• 104 PD patients were assessed using an optoelectronic system for gait analysis.
• Data collected during single and dual tasks.
• Statistical analysis followed by machine learning (Decision Tree, Support Vector Machine) for classification.

Main Results:

• PD patients with NPS exhibited slower gait, increased instability, and reduced movement control.
• Machine learning models achieved high accuracy (up to 91.3%) using kinematic and kinetic features.
• Classifiers identified gait features consistent with statistical analysis findings.

Conclusions:

• Gait analysis is a valuable, non-invasive biomarker for identifying PD patients with a higher burden of NPS.
• Machine learning models trained on gait data can accurately distinguish between PD patients with and without NPS.
• This approach supports early diagnosis, personalized monitoring, and integration into clinical workflows.