Locomotion can be investigated by factorization of electromyographic (EMG) signals, e.g. with non-negative matrix factorization (NMF). This approach is a convenient concise representation of muscle activities as distributed in motor modules, activated in specific gait phases. For applying NMF, the EMG signals are analysed either as single trials, or as averaged EMG, or as concatenated EMG (data structure). The aim of this study is to investigate the influence of the data structure on the extracted motor modules. Twelve healthy men walked at their preferred speed on a treadmill while surface EMG signals were recorded for 60 s from 10 lower limb muscles. Motor modules representing relative weightings of synergistic muscle activations were extracted by NMF from 40 step cycles separately (EMGSNG), from averaging 2, 3, 5, 10, 20 and 40 consecutive cycles (EMGAVR), and from the concatenation of the same sets of consecutive cycles (EMGCNC). Five motor modules were sufficient to reconstruct the original EMG datasets (reconstruction quality > 90%), regardless of the type of data structure used. However, EMGCNC was associated with a slightly reduced reconstruction quality with respect to EMGAVR. Most motor modules were similar when extracted from different data structures (similarity > 0.85). However, the quality of the reconstructed 40-step EMGCNC datasets when using the muscle weightings from EMGAVR was low (reconstruction quality ~ 40%). On the other hand, the use of weightings from EMGCNC for reconstructing this long period of locomotion provided higher quality, especially using 20 concatenated steps (reconstruction quality ~ 80%). Although EMGSNG and EMGAVR showed a higher reconstruction quality for short signal intervals, these data structures did not account for step-to-step variability. The results of this study provide practical guidelines on the methodological aspects of synergistic muscle activation extraction from EMG during locomotion.

Balance recovery during walking requires complex sensory-motor integration. Mechanisms to avoid falls are active concomitantly with human locomotion motor patterns. It has been suggested that gait can be described by a set of motor modules (synergies), but little is known on the modularity of gait during recovery of balance due to unexpected slips. Our hypothesis was that muscular activation during reactive recovery of balance during gait has a modular organization. The aim of the study was to verify this hypothesis when perturbations were delivered in different directions. Eight healthy men walked on a 7-m walkway, which had a moveable force platform embedded in the middle. Subjects experienced unperturbed walking as well as perturbations delivered in the sagittal (forward and backward) and frontal (leftward and rightward) planes. Bilateral full-body kinematics and surface electromyography (EMG) from lower limbs, trunk, and neck were recorded during walking. Synergies and activation signals were extracted from surface EMG signals. Four modules were sufficient to explain the unperturbed gait and the gait perturbed in any of the perturbation directions. Moreover, three of four modules extracted from the unperturbed gait were the same for gait perturbed forward, leftward, and rightward (similarity in synergies = 0.94 ± 0.03). On the other hand, the activation signals were different between unperturbed and perturbed gait (average correlation coefficient = 0.55 ± 0.16). These strategies to recover balance were robust across subjects. In conclusion, changes in lower limb and trunk kinematics provoked by perturbations were reflected in minimal adjustments in the muscular modular organization of walking, with three of four modules preserved from normal walking. Conversely, the activation signals were all substantially influenced by the perturbations, being the result of integration of afferent information and supraspinal control.

The mechanics of cutting movements have been investigated extensively, but few studies have considered the rapid deceleration phase prior to turning which has been linked to muscle damage. This study used accelerometry to examine the influence of turning intensity on the last three steps of a severe turn. Ten soccer players performed 135° “V” cuts at five different intensities. Resultant decelerations were recorded from a trunk-mounted tri-axial accelerometer. Lower limb kinematics and ground reaction forces (GRF) from the pivot foot-ground contact (FGC) were also monitored. Average peak trunk decelerations were larger at the two preceding steps (4.37 ± 0.12 g and 4.58 ± 0.11 g) compared to the PIVOT step (4.10 ± 0.09 g). Larger peak joint flexion angular velocities were observed at PRE step (ankle: 367 ± 192 deg.s⁻¹; knee 493 ± 252 deg.s⁻¹) compared to PIVOT step (ankle 255 ± 183 deg.s⁻¹; knee 377 ± 229 deg.s⁻¹). Turn intensity did not influence peak GRF at PIVOT step. This study highlights the importance of steps prior to turning and their high-frequency loading characteristics. It is suggested that investigations of lower limb loading during turning should include this deceleration phase and not focus solely on pivot FGC.

Motorized treadmills have been widely used in locomotion studies, although a debate remains concerning the extrapolation of results obtained from treadmill experiments to overground locomotion. Slight differences between treadmill (TRD) and overground running (OVG) kinematics and muscle activity have previously been reported. However, little is known about differences in the modular control of muscle activation in these two conditions. Therefore, we aimed at investigating differences between motor modules extracted from TRD and OVG by factorization of multi-muscle electromyographic (EMG) signals. Twelve healthy men ran on a treadmill and overground at their preferred speed while we recorded tibial acceleration and surface EMG from 11 ipsilateral lower limb muscles. We extracted motor modules representing relative weightings of synergistic muscle activations by non-negative matrix factorization from 20 consecutive gait cycles. Four motor modules were sufficient to accurately reconstruct the EMG signals in both TRD and OVG (average reconstruction quality = 92±3%). Furthermore, a good reconstruction quality (80±7%) was obtained also when muscle weightings of one condition (either OVG or TRD) were used to reconstruct the EMG data from the other condition. The peak amplitudes of activation signals showed a similar timing (pattern) across conditions. The magnitude of peak activation for the module related to initial contact was significantly greater for OVG, whereas peak activation for modules related to leg swing and preparation to landing were greater for TRD. We conclude that TRD and OVG share similar muscle weightings throughout motion. In addition, modular control for TRD and OVG is achieved with minimal temporal adjustments, which were dependent on the phase of the running cycle.