Systematic characterisation of silicon-embedded accelerometers for mechanomyography

Area of Science:

• Biomedical Engineering
• Prosthetics and Orthotics
• Sensor Technology

Background:

• Current externally powered prostheses for below-elbow amputees face challenges with reliable sensor placement.
• Electromyography (EMG) sensors are difficult to secure within silicon prosthetic sockets, hindering functionality and comfort.
• Mechanomyography (MMG) offers an alternative by measuring muscle vibrations, with sensors not requiring direct skin contact.

Purpose of the Study:

• To determine the feasibility of recording mechanomyography (MMG) signals using silicon-embedded, micro-machined accelerometers.
• To assess the performance of embedded MMG sensors compared to non-embedded ones for potential use in prosthetic control.

Main Methods:

• Fifteen micro-machined accelerometers were embedded within silicon material.
• Embedded and non-embedded accelerometers were subjected to predefined vibration patterns.
• Signal-to-noise ratio (SNR) and frequency response were measured and compared between embedded and non-embedded samples.

Main Results:

• Silicon-embedded accelerometers demonstrated a significantly higher signal-to-noise ratio (approximately 19 dB) compared to non-embedded ones (approximately 12 dB).
• A considerable mechanical damping effect of silicon was observed in the 300-900 Hz bandwidth, enhancing SNR (p=0.0028).

Conclusions:

• Silicon-embedded accelerometers are feasible for recording MMG signals.
• This technology offers a promising solution for reliable sensor integration in externally powered prostheses.
• Improved MMG sensing could enhance the functionality and user experience of prosthetic devices.