A motion-tolerant adaptive algorithm for wearable photoplethysmographic biosensors

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Signal Processing

Background:

• Motion artifact significantly degrades the performance of wearable biosensors.
• Accurate extraction of physiological parameters like heart rate (HR) and blood oxygen saturation (SpO2) is crucial for remote health monitoring.
• Photoplethysmographic (PPG) signals are susceptible to motion-induced noise, limiting the utility of many wearable devices.

Purpose of the Study:

• To develop and validate a novel real-time adaptive algorithm for motion-tolerant extraction of HR and SpO2 from PPG signals.
• To effectively remove motion artifacts originating from various sources, including tissue and venous blood changes.
• To provide noise-free PPG waveforms for enhanced feature extraction in wearable biosensors.

Main Methods:

• A two-stage normalized least mean square adaptive noise canceler was designed.
• A novel synthetic reference signal was utilized for algorithm validation at each stage.
• The algorithm's performance was evaluated using Bland-Altman agreement and correlation analyses against commercial ECG and SpO2 sensors.

Main Results:

• The algorithm demonstrated high agreement and correlation for HR ( > 0.98) and SpO2 ( > 0.7) extraction.
• Successful noise-free PPG waveform extraction was achieved during standing, walking, and running.
• Validation was performed across single- and multi-subject scenarios under various conditions.

Conclusions:

• The proposed real-time adaptive algorithm effectively mitigates motion artifacts in wearable PPG biosensors.
• Accurate and reliable extraction of HR and SpO2 is achievable even during physical activity.
• This technology enhances the performance and reliability of wearable health monitoring devices.