G M Eom | PubMed Health

PubMed2002

Gradual potentiation of isometric muscle force during constant electrical stimulation.

An investigation was carried out into how stimulation frequency and stimulation history affect the potentiation of muscle force during 20s of constant stimulation of the two knee extensors in isometric conditions. Stimulation frequency significantly affected the potentiation pattern: low-frequency (2.5-10 Hz) stimulation showed a reduction and subsequent enhancement of force, and high-frequency (14.3-25 Hz) stimulation showed only enhancement of force. The degree of enhancement in force and time-to-peak decreased with the stimulation frequency. Whereas 40 Hz conditioning stimulation enhanced the muscle force, 14 Hz stimulation after 10s of rest induced little force enhancement (8% in both muscles). When the frequency of the conditioning stimulation was 14 Hz and the main 14 Hz stimulation was applied after 50s of rest, the initial force at the main stimulation was similar to the final force value of the conditioning stimulation (above 90% similarity). The potentiated twitch force slowly decayed during rest, with an average time constant of 2.4 min. These observations indicate that muscle potentiation depends on the stimulation frequency and stimulation history, and therefore a computer model of potentiation can play an important role in predicting muscle force and body movement induced by electrical stimulation.

Adult↗

PubMed2000

Computer-aided generation of stimulation data and model identification for functional electrical stimulation (FES) control of lower extremities.

Standard stimulation data for unassisted standing up of paraplegic patients was generated by dynamic optimization linked with model simulation, to overcome the difficulties in the present electromyogram (EMG)-based method. The generated stimulation data were roughly in agreement with the normal subjects' EMG. From these, it is suggested that the 'model-based' method is useful as an alternative of the 'EMG-based method'. The same technique can be applied to generation of patient-specific stimulation data once the musculoskeletal system of a patient is properly identified. The musculoskeletal system must be identified from data taken from simple and noninvasive experiments for the identification method to be practically acceptable. We developed a musculoskeletal model and systematic identification protocols for this purpose. They were validated for the vastus lateralis muscle at the knee joint. The identification was successful and the predicted joint angle trajectories closely matched the experimental data. This implies that the model-based generation of patient-specific stimulation data is possible.

Computer Simulation↗