Rest tremor quantification based on fuzzy inference systems and wearable sensors

Area of Science:

• Biomedical Engineering
• Neurology
• Wearable Technology

Background:

• The Movement Disorder Society sponsored Unified Parkinson Disease Rating Scale (MDS-UPDRS) is the standard for rating Parkinson's disease (PD).
• Current MDS-UPDRS motor assessments rely on subjective human interpretation.
• Existing wearable sensor applications for PD analysis often deviate from MDS-UPDRS standards, lacking clear feature-to-rating connections.

Purpose of the Study:

• To develop a novel method for objective rest tremor quantification in Parkinson's disease (PD).
• To create a system that adheres to the MDS-UPDRS standards using wearable sensors and a fuzzy inference system.
• To enhance the reliability, repeatability, and resolution of tremor assessments in clinical settings.

Main Methods:

• Fifty-seven PD patients were assessed across twelve sessions, yielding 123 measurements.
• Limb features correlated to rest tremor amplitude were extracted using gyroscopes, accelerometers, and magnetometers.
• A fuzzy inference system processed 446 feature combinations to estimate tremor severity.

Main Results:

• A novel method for rest tremor quantification, fully aligned with MDS-UPDRS, was successfully developed.
• The system provides reliable, repeatable assessments and computes clinician-suggested features.
• The method offers immediate clinical accessibility, provides timeline-based rest tremor amplitude data, and improves rating resolution with a continuous range.

Conclusions:

• The proposed wearable sensor and fuzzy inference system offer a standardized, objective, and scalable approach to PD rest tremor assessment.
• This method enhances diagnostic capabilities by providing continuous, high-resolution tremor data.
• The system's accessibility and adherence to MDS-UPDRS standards facilitate improved clinical inference and patient monitoring.