PPGSynth
PPGSynth synthesizes regular and arrhythmic photoplethysmography (PPG) waveforms to generate large synthetic datasets for research and validation of cardiovascular monitoring and arrhythmia analysis algorithms.
Key Features:
- Synthesis of Regular and Irregular Heartbeats: Generates PPG waveforms with regular beats and three types of irregular heartbeats: compensation, interpolation, and reset.
- Modeling Approach: Models each PPG pulse as two combined Gaussian functions to approximate physiological waveform morphology.
- Customization Options: Provides adjustable parameters including irregularity level, sampling frequency, and waveform duration.
- Noise Modeling: Adds configurable noise types, specifically Gaussian noise and multi-frequency noise, to simulate real-world signal conditions.
- Robust Testing Capabilities: Produces large, configurable synthetic datasets for testing and validating algorithms that analyze arrhythmic PPG signals.
Scientific Applications:
- Algorithm Development and Validation: Enables creation of extensive labeled datasets for developing and benchmarking arrhythmia detection and PPG-analysis algorithms.
- Research in Cardiovascular Diseases: Provides realistic synthetic PPG signals for studying physiological signatures relevant to cardiovascular diagnostics.
- Digital Health Evaluation: Supports testing of devices and signal-processing software against a range of simulated physiological and noise scenarios.
Methodology:
Each PPG pulse is modeled using two combined Gaussian functions; the generator synthesizes regular and three irregular heartbeat types (compensation, interpolation, reset) with adjustable irregularity level, sampling frequency, waveform duration, and optional Gaussian or multi-frequency noise.
Topics
Details
- Added:
- 1/18/2021
- Last Updated:
- 1/27/2021
Operations
Publications
Tang Q, Chen Z, Allen J, Alian A, Menon C, Ward R, Elgendi M. PPGSynth: An Innovative Toolbox for Synthesizing Regular and Irregular Photoplethysmography Waveforms. Frontiers in Medicine. 2020;7. doi:10.3389/fmed.2020.597774. PMID:33224967. PMCID:PMC7668389.