paradigma

Paradigma processes and extracts features from wearable sensor data to identify digital biomarkers specifically tailored for Parkinson's disease.

Key Features:

• Multi-signal processing: The toolbox processes accelerometer, gyroscope and photoplethysmography signals collected during passive monitoring in daily life.
• Validated pipelines: It contains three scientifically validated data processing pipelines: arm swing during gait, tremor, and pulse rate analysis.
• Orchestrator function: An orchestrator function enables end-to-end processing from raw data loading to aggregated measures, with automatic data preparation including flexible column mapping and sensor orientation adjustment.
• General signal processing functionalities: The toolbox provides general functionalities for signal processing and feature extraction, such as filtering, peak detection, and spectral analysis.
• Modular architecture: The modular architecture enables researchers to extend the toolbox with custom algorithms and functionalities while maintaining compatibility with standardized data formats.

Scientific Applications:

• Parkinson's disease digital marker derivation: The toolbox extracts digital measures of motor and non-motor signs of Parkinson's disease from real-life wrist sensor data.
• Arm swing analysis during gait: The gait pipeline extracts digital markers related to arm swing during walking.
• Tremor quantification: The tremor pipeline extracts digital markers of tremor from continuous wearable sensor data.
• Pulse rate analysis: The pulse rate analysis pipeline extracts cardiac-related digital markers from photoplethysmography signals.

Methodology:

The toolbox implements three scientifically validated data processing pipelines: arm swing during gait, tremor, and pulse rate analysis. An orchestrator function handles end-to-end processing from raw data loading to aggregated measures, with automatic data preparation including flexible column mapping and sensor orientation adjustment. General signal processing methods include filtering, peak detection, and spectral analysis.