Adaptive Auto-Regressive Proportional Myoelectric Control.

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Proportional myoelectric control allows users to regulate movement position or velocity.
• Existing methods like moving-average filters have limitations in adaptability and performance.
• Adaptive control strategies are needed to optimize human-machine interaction.

Purpose of the Study:

• To introduce adaptive auto-regressive filters for seamless adjustment between position and velocity control in myoelectric interfaces.
• To evaluate the performance of this adaptive system in a closed-loop, human-machine cursor control task.
• To compare the effectiveness of auto-regressive filters against previous moving-average filter approaches.

Main Methods:

• Implementation of an adaptive system with closed-loop feedback for simultaneous user and machine cursor control.
• Utilizing an eight-channel myoelectric armband for data acquisition from participants.
• Testing the system with 15 able-bodied and 3 limb-deficient individuals.

Main Results:

• Human-machine pairs demonstrated smoother cursor movements and a larger range of motion with auto-regressive filters compared to moving-average filters.
• The system converged towards a velocity control strategy, leading to faster and more accurate movements.
• Reduced muscle effort was observed when using the adaptive auto-regressive filters.

Conclusions:

• Adaptive auto-regressive filters offer a superior method for proportional myoelectric control, enhancing movement smoothness, accuracy, and efficiency.
• The developed human-machine system effectively adapts to optimize cursor control, reducing user effort.
• This adaptive control approach is versatile and applicable to various high-dimensional signal-based motion control applications, including brain-computer interfaces.