Identification of spectral features within muscle synergies during the generation of isometric forces

The human body, a mechanically redundant structure, relies on the intricate interplay of the skeleton, muscles, and ligaments to perform precise motor tasks. Central to understanding motor control is the coordination among these components, orchestrated primarily by the Central Nervous System (CNS). Recent research highlights that muscle activity is governed by modular structures known as muscle synergies, which simplify control mechanisms by reducing computational load. Recently, the present group of researchers has developed a novel method for quantifying muscle synergies in time frequency-domain, using surface electromyographic (sEMG) signals and Continuous Wavelet Transform (CWT). This method, named the pooled scalogram, has been tested in gait analysis but shows promise for other motor tasks. The present study aims to evaluate the pooled scalogram’s efficacy in quantifying time-frequency synergies among multiple muscles during an upper limb isometric motor task. The process involved four main phases: 1) applying Non-negative Matrix Factorization (NMF) to sEMG signals from 17 muscles; 2) identifying active muscles in detected synergies; 3) calculating CWT coefficients and scalograms for these muscles; 4) calculating pooled scalogram through CWT coefficients matrices, in order to evaluate synergies in time-frequency domain. The present results suggested that the pooled scalogram is able to characterize muscle correlations and quantify synergies, offering deeper insights in both time and frequency domains compared to existing methods. This work provides a foundation for future applications in various clinical and sports settings, where understanding muscle coordination is crucial.