Coherence between Decomposed Components of Wrist and Finger PPG Signals by Imputing Missing Features and Resolving...

Area of Science:

• Biomedical Engineering
• Physiological Signal Processing

Background:

• Photoplethysmography (PPG) signal analysis is crucial for hemodynamic and vascular insights.
• Variations in PPG waveform morphology across measurement sites (e.g., finger, wrist) can lead to missing or ambiguous features.
• These signal imperfections hinder accurate feature extraction and subsequent analysis.

Purpose of the Study:

• To analyze the causes of missing or ambiguous features in finger and wrist PPG signals.
• To propose a systematic approach for imputing missing features and resolving ambiguous ones.
• To enhance the reliability and availability of PPG features for further processing.

Main Methods:

• Analysis of PPG pulse decomposition into component waves to understand feature ambiguities.
• Development and application of a systematic imputation and resolution technique for PPG signal features.
• Experimental validation using large datasets of finger and wrist PPG cycles.

Main Results:

• Successful identification of features in over 98.7% of finger and 99.1% of wrist PPG cycles.
• Increased stability and reduced standard deviation in extracted PPG features.
• High correlation (up to 0.92) observed between finger and wrist PPG waveforms in component wave characteristics.

Conclusions:

• The proposed method effectively addresses missing and ambiguous PPG features, expanding feature availability.
• Enhanced feature extraction reveals deeper intrinsic properties of finger and wrist PPG signals.
• Demonstrated coherence between finger and wrist PPG enhances the utility of wrist-based PPG measurements.