Human Neuromuscular System Identification Using Functional Electrical Stimulation for the Development of a Digital...

Area of Science:

• Biomechanics and Motor Control
• Human Digital Twin Technology
• Neuromusculoskeletal System Modeling

Background:

• Human digital twin development for the locomotor system is ongoing.
• A method using functional electrical stimulation (FES) estimates neuromusculoskeletal system dynamics.
• This method uses electrical agonist-antagonist ratio (rE) and sum (sE) of stimulation intensities.
• Previous work showed neuromuscular system (NMS) parameters vary with sE but this variation was unmodeled.

Purpose of the Study:

• To investigate the influence of sE on isometric elbow joint motion parameters.
• To model the relationship between sE and key NMS parameters.
• To evaluate different mathematical models for describing these relationships.

Main Methods:

• Experiments were conducted under isometric contraction with 10 participants.
• Parameters of a second-order system (Kp, ωn, ζ) were estimated at 15 different sE levels.
• Model performance was evaluated using the corrected Akaike information criterion (AICc) for linear, quadratic, and exponential models.

Main Results:

• For proportional gain (Kp), a concave quadratic model best fit group mean data.
• For natural frequency (ωn) and damping ratio (ζ), convex quadratic models best described group mean data.
• Individual data showed some participants exhibited monotonic trends, and the exponential model showed comparable AICc values for Kp.

Conclusions:

• The study provides models for how sE influences NMS parameters, essential for advancing human digital twins.
• These models can potentially enable real-time estimation of human movement from electromyographic (EMG) signals.
• Further research is needed to explore parameter trends at stimulation frequencies closer to EMG signals, considering the refractory period effect.