A Method for Identification of Mechanical Response of Motor Units in Skeletal Muscle Voluntary Contractions Using Ultrafast Ultrasound Imaging—Simulations and Experimental Tests

Abstract

The central nervous system coordinates movement through forces generated by motor units (MUs) in skeletal muscles. To analyze MUs function is essential in sports, rehabilitation medicine applications, and neuromuscular diagnostics. The MUs and their function are studied using electromyography. Typically, these methods study only a small muscle volume (1 mm 3) or only a superficial (<1 cm) volume of the muscle. Here we introduce a method to identify so-called mechanical units, i.e., the mechanical response of electrically active MUs, in the whole muscle (4 × 4 cm, cross-sectional) under voluntary contractions by ultrafast ultrasound imaging and spatiotemporal decomposition. We evaluate the performance of the method by simulation of active MUs' mechanical response under weak contractions. We further test the experimental feasibility on eight healthy subjects. We show the existence of mechanical units that contribute to the tissue dynamics in the biceps brachii at low force levels and that these units are similar to MUs described by electromyography with respect to the number of units, territory sizes, and firing rates. This study introduces a new potential neuromuscular functional imaging method, which could be used to study a variety of questions on muscle physiology that previously were difficult or not possible to address.

A Deep Learning Pipeline for Identification of Motor Units in Musculoskeletal Ultrasound

“…2) Unknown number of units: The number of active (and thus visible) MUs in an image sequence is unknown because the recruitment of units is highly complex coordination by the central nervous system [3], [10]. 3) Overlapping territories of the units: The territories of two or more MUs may be overlapping [3], causing spatial and temporal interference of their activity ( Fig 1A and C).…”

“…We evaluate the extracted signals' performance by comparing the firing patterns of the estimated and simulated signals. The performance metric selected for this task is the rate of agreement (RoA), as it provides easy interpretation, captures the information we are interested in [35], and was used in previous work on similar signals [10]. The RoA measure ranges between 0 and 1 where 1 is a perfect score, and it measures the agreement between the firings of two signals.…”

“…A simulator [10] generated the data used for training and evaluating the models in this work. Since the simulator knows all the latent variables used to generate the data, it can also provide the labels for the example in the form of masks of the cross-sectional territory encompassed by each MU and the mechanical twitch train signal for each MU.…”

“…Recently our group presented a method [10] to study MU activity based on the mechanical response of individual MUs using ultrafast ultrasound imaging (>2000 images per second [11]). This method was based on a blind source separation framework and decomposition of spatio-temporal components.…”

“…This work aims to develop and evaluate a deep learning pipeline to 1) detect individual active MUs, 2) segment their territories, and 3) estimate their activation twitch signal, using ultrafast image sequences of voluntary skeletal muscle contractions. The model is trained and evaluated using simulated ultrasound data [10].…”