PPG2EMG: Estimating Upper-Arm Muscle Activities and EMG from Wrist PPG Values

Area of Science:

• Biomedical Engineering
• Signal Processing
• Human-Computer Interaction

Background:

• Electromyogram (EMG) measures muscle electrical activity but surface EMG (sEMG) requires costly consumables and can irritate skin.
• Existing research focuses on EMG for human activity recognition, particularly for arm and hand function support.
• There is a need for non-invasive, cost-effective methods for monitoring muscle activity.

Purpose of the Study:

• To propose a novel method for recognizing arm muscle activity states using photoplethysmography (PPG) data.
• To develop a method for estimating EMG signals from PPG data.
• To explore PPG as a viable alternative to traditional EMG for muscle activity monitoring.

Main Methods:

• Utilized PPG sensors worn on the wrist to measure pulse wave data during arm muscle exertion.
• Correlated changes in pulse wave characteristics (e.g., amplitude reduction) with upper arm muscle contraction.
• Developed algorithms to recognize muscle activity states and estimate EMG based on PPG signal variations.

Main Results:

• Successfully recognized three types of arm muscle activity with over 80% accuracy.
• Estimated EMG signals from PPG data with an approximate 20% error rate.
• Demonstrated the potential of PPG for non-invasive muscle activity assessment.

Conclusions:

• PPG-based monitoring offers a promising, cost-effective approach for arm muscle activity recognition and EMG estimation.
• This method could reduce reliance on traditional sEMG sensors, mitigating associated costs and skin irritation.
• Further research can refine PPG-based muscle monitoring for broader applications in rehabilitation and human-computer interaction.