A model of the surface electromyogram in pathological tremor.

Abstract

The study developed a novel multi-scale model for simulating the surface electromyogram (EMG) of an antagonistic pair of muscles during pathological tremor. By combining and expanding mathematical descriptions from motor units to limb kinematics, the model constitutes the first attempt to simulate the surface EMG and the individual motor unit activity under the influence of descending voluntary command, oscillatory noise in the descending signal, and afferent feedback when controlling a freely moving limb to achieve a predefined angular trajectory. The oscillatory noise was adjusted to simulate various types of pathological tremor. The simulations replicated previously reported experimental results for the power spectral density of the surface EMG, the angular velocity of the limb, and single motor unit activity. The model provides a powerful tool for extracting information about how the surface EMG can be used to describe tremor in various conditions, including different tremor frequencies and intensities, that cannot be achieved solely with experimental approaches.