Validation of an artificially activated mechanistic muscle model by using inverse dynamics analysis

• An approach to estimate FES profiles from a given movement is presented.
• FES profiles can be obtained in terms of pulse-width, frequency or amplitude.
• The present proposal takes the physiology of the process into account.
• Applied FES and profiles obtained through the model show acceptable correlation.
• This method can be used to design subject-specific FES rehabilitation programs.

The objective of this work is to validate a mechanistic muscle model by using inverse dynamics analysis. To do so, an artificially activated Hill-type muscle model is used. Compared to the traditional physiologically activated muscle model, an artificially activated model must take into account an additional set of parameters and dynamics that can affect the resulting force. To validate the model, the ankle dorsiflexion activated by functional electrical stimulation (FES) is subjected to an inverse dynamic analysis (IDA). The resulting values of the net joint torques are used to estimate first the muscle forces, and second, by inversion of the proposed artificially activated model, the stimulation profile that produces the recorded motion. The results are then compared with the stimulation profile applied to the subject, and the model parameters are adjusted correspondingly. Once the model has been validated, the methodology could be used to design rehabilitation programs based on electrical stimulation or to prescribe FES actuation in the design of hybrid orthoses or neuroprostheses to achieve a given movement.