A Systematic Review of Sensor Fusion Methods Using Peripheral Bio-Signals for Human Intention Decoding.

Area of Science:

• Biomedical Engineering
• Human-Computer Interaction
• Rehabilitation Engineering

Background:

• The increasing use of robotic and prosthetic devices necessitates intuitive interfaces for embodied human-device interaction.
• Muscle-machine interfaces (MuMIs) offer a solution by decoding user intentions through myoelectric signals.
• Various myography techniques exist, each with unique advantages and limitations.

Purpose of the Study:

• To systematically review and analyze the advantages and disadvantages of different myography methods for MuMIs.
• To evaluate the effectiveness of sensor fusion techniques in decoding user intentions.
• To identify promising sensor fusion strategies for various applications, considering interface wearability.

Main Methods:

• A systematic review adhering to PRISMA guidelines was conducted.
• Studies employing the fusion of different sensors and myography techniques were identified and analyzed.
• Wearability and the properties of different fusion techniques in decoding user intentions were explored.

Main Results:

• The fusion of two or more myography methods demonstrably improves the performance in decoding user intentions.
• Electromyography, ultrasonography, mechanomyography, and near-infrared spectroscopy are key modalities explored in sensor fusion.
• Sensor fusion techniques, including inertial measurement units and optical sensing, show significant interest for upper limb intention decoding.

Conclusions:

• Sensor fusion is crucial for enhancing the performance of muscle-machine interfaces.
• Combining multiple myography techniques offers superior user intention decoding compared to single methods.
• This review identifies effective sensor fusion strategies for developing advanced MuMIs across diverse applications.