Physical and electrophysiological motor unit characteristics are revealed with simultaneous high-density electromyography and ultrafast ultrasound imaging

Abstract

Electromyography and ultrasonography provide complementary information about electrophysiological and physical (i.e. anatomical and mechanical) muscle properties. In this study, we propose a method to assess the electrical and physical properties of single motor units (MUs) by combining High-Density surface Electromyography (HDsEMG) and ultrafast ultrasonography (US). Individual MU firings extracted from HDsEMG were used to identify the corresponding region of muscle tissue displacement in US videos. The time evolution of the tissue velocity in the identified region was regarded as the MU tissue displacement velocity. The method was tested in simulated conditions and applied to experimental signals to study the local association between the amplitude distribution of single MU action potentials and the identified displacement area. We were able to identify the location of simulated MUs in the muscle cross-section within a 2 mm error and to reconstruct the simulated MU displacement velocity (cc > 0.85). Multiple regression analysis of 180 experimental MUs detected during isometric contractions of the biceps brachii revealed a significant association between the identified location of MU displacement areas and the centroid of the EMG amplitude distribution. The proposed approach has the potential to enable non-invasive assessment of the electrical, anatomical, and mechanical properties of single MUs in voluntary contractions.

Kinematics of individual muscle units in natural contractions measured in vivo using ultrafast ultrasound

“…Further limitations include the use of systems with low temporal resolution compared to the durations of the involved physiological processes (Botter et al ., 2013), low spatial resolution to properly separate contributions from multiple MUs (Birkbeck et al ., 2020), or the deployment of invasive techniques, impacting the mechanics of the system and enabling the detection of just one MU at a time (Rohlén, Stålberg and Grönlund, 2020). The use of ultrafast US to decompose contributions to muscle motion from individual MUs has been previously attempted using spatio-temporal independent component analysis (Rohlén et al ., 2020; Rohlén, Stålberg and Grönlund, 2020; Carbonaro et al ., 2022), whilst we propose an approach which isolates these contributions by directly using the known MU discharge times.…”

“…were not identified in a study using ultrafast US (Carbonaro et al, 2022), presumably because of methodological constraints. Our work therefore presents the first method able to detect both split and overlapping regions of activity.…”

“…US has also been used for studying the mechanical response of individual MUs in electrically stimulated contractions (Deffieux et al, 2008;Waasdorp et al, 2019Waasdorp et al, , 2021. Furthermore, some studies have suggested that US can even be used at the muscle unit level (Rohlén et al, 2020;Rohlén, Stålberg and Grönlund, 2020;Carbonaro et al, 2022).…”

“…US has also been used for studying the mechanical response of individual MUs in electrically stimulated contractions (Deffieux et al ., 2008; Waasdorp et al ., 2019, 2021). Furthermore, some studies have suggested that US can even be used at the muscle unit level (Rohlén et al ., 2020; Rohlén, Stålberg and Grönlund, 2020; Carbonaro et al ., 2022). Here, we utilise US alongside a new signal processing pipeline, with minimal processing and assumptions, to extract detailed new characteristics of the kinematics of single muscle units.…”

Estimating the neural spike train from an unfused tetanic signal of low threshold motor units using convolutive blind source separation

“…Ultrafast ultrasound has been shown to image and analyse voluntarily activated MUs for a large field of view in the muscle (40x40 mm) [3][4][5][6][7][8] , providing spatiotemporal mechanics at a high resolution (<1 mm and >1 kHz). This technique is based on recording radio frequency signals (B-mode images) (Fig.…”