EMG and metabolite-based prediction of force in paralyzed quadriceps muscle under interrupted stimulation

Abstract

A major issue associated with functional electrical stimulation (FES) of a paralyzed limb is the decay with time of the muscle force as a result of fatigue. A possible means to reduce fatigue during FES is by using interrupted stimulation, in which fatigue and recovery occur in sequence. In this study, we present a model which enables us to evaluate the temporal force generation capacity within the electrically activated muscle during first stimulation fatigue, i.e., when the muscle is activated from unfatigued initial conditions, and during postrest stimulation, i.e., after different given rest durations. The force history of the muscle is determined by the activation as derived from actually measured electromyogram (EMG) data, and by the metabolic fatigue function expressing the temporal changes of muscle metabolites, from existing data acquired by in vivo 31P MR spectroscopy in terms of the inorganic phosphorus variables, Pi or H2PO4-, and by the intracellular pH. The model was solved for supra-maximal stimulation in isometric contractions separated by rest periods, and compared to experimentally obtained measurements. EMG data were fundamental for prediction of the ascending force during its posttetanic response. On the other hand, prediction of the decaying phase of the force was possible only by means of the metabolite-based fatigue function. The prediction capability of the model was assessed by means of the error between predicted and measured force profiles. The predicted force obtained from the model in first stimulation fatigue fits well with the experimental one. In postrest stimulation fatigue, the different metabolites provided different prediction capabilities of the force, depending on the duration of the rest period. Following rest duration of 1 min, Pi provided the best prediction of force; H2PO4- extended the prediction capacity of the model to up to 6 min and pH provided a reliable prediction for rest durations longer than 12 min. The results presented shed light on the roles of EMG and of metabolites in prediction of the force history of a paralyzed muscle under conditions where fatigue and recovery occur in sequence.