The Use of Electromyography for the Noninvasive Prediction of Muscle Forces

SummarySuitably processed electromyographic (EMG) signals can be combined with Hill-type musculoskeletal models to noninvasively achieve estimations of individual muscle forces. This method has particular advantages over other methods for the assessment of a given performance. The purpose of this review is to report on the current issues facing the human movement scientist who wishes to extend the kinetic information yielded by linked segment models to the kinetics of individual muscles. Such an extension is necessary when considering co-contraction of antagonistic muscles, the role of bi-articular muscles, coordination, movement efficiency or bone-on-bone forces.Currently, linked segment models have not been successfully extended to individual muscle forces for diagnostic purposes by using the EMG approach or any other approach. Most models have been designed for a specific purpose and have only been evaluated over a narrow range of movement conditions. More generalised models will require greater complexity and possibly more extensive calibration or an increased number of specific inputs or greater computational effort. This review shows the promise of the EMG approach and presents the challenges, as well as the strategies, that should enable more general, accurate and precise estimates of individual muscle forces.