Exploring surface electromyography (EMG) as a feedback variable for the human-in-the-loop optimization of lower limb...

Area of Science:

• Biomechanics
• Robotics
• Human-Computer Interaction

Background:

• Human-in-the-loop (HITL) optimization aims to reduce walking effort in wearable robotics, often using metabolic cost feedback.
• Metabolic cost feedback is time-intensive, necessitating alternative, more efficient feedback variables.

Purpose of the Study:

• To investigate lower limb surface electromyography (EMG) as a potential alternative feedback variable for HITL optimization.
• To determine if EMG can reliably distinguish between different walking effort conditions.

Main Methods:

• A laboratory experiment involved 13 subjects walking on a treadmill with varying lower limb weight loads (2, 4, 8 kg).
• Surface EMG of seven lower limb muscles was recorded, and mean absolute values per stride were analyzed.
• Detection rates for changes in walking effort were assessed using single/multiple strides and single/multiple muscles, alongside EMG drift evaluation.

Main Results:

• Combining multiple consecutive strides and multiple muscles significantly increased EMG's detection rate for walking effort changes.
• EMG drift was most pronounced during warmup and at the start of weight sessions.
• Reliable EMG feedback for HITL optimization requires approximately 10 consecutive strides (5.5s) and 16.5s of walking post-acclimatization.

Conclusions:

• Surface EMG is a promising alternative feedback variable for HITL optimization in lower limb wearable robotics.
• Optimal EMG-based feedback requires combining data from multiple muscles and strides, with adequate warmup and minimal breaks to mitigate drift.
• Future research should focus on refining EMG-based feedback variables and reducing their variability for enhanced HITL optimization.