Trigno system

Surface EMG was measured using the Delsys Trigno system (Delsys Inc., Boston, MA, USA), which consists of portable wireless EMG sensors with four bar electrodes that are attached to the skin directly above the muscle to be examined. Surface EMG sensors record the summated action potentials produced by the muscle fibers of a contracting muscle. Each sensor recorded 3-axis gyroscope and accelerometer data and time data in synchrony with the EMG data.
Four EMG channels were used, each designated to one of the four lower-limb muscles of the more affected limb (VL, BF, TA and medial GAST) according to SENIAM guidelines. An additional sensor was placed at the medial malleolus of the examined limb to record the accelerometer data for gait phase detection. The EMG channels were synchronized in time, facilitating analysis with reference to the gait phases.

Forces generated by the participants were recorded at a handle instrumented with a 6-axis force-torque transducer (Omega160, ATI Industrial Automation, Apex, NC, United States) (Figure 1A). Force was sampled at 120 Hz using custom software based on Dragonfly acquisition system (Pittsburgh, 2017 ).
Surface EMG signals were recorded with a wireless Trigno System (Delsys Inc., Boston, MA, United States). EMG activity was recorded from 14 major muscles of the shoulder, arm and forearm of the participant’s dominant upper limb: superior (ST) and middle trapezius (MT), infraspinatus (Inf), serratus anterior (SA), anterior (AD), middle (MD), and posterior deltoid (PD), pectoralis major (PM, clavicular fibres), short (BS) and long (BL) head of biceps brachii, long (TL) and lateral (TLat) heads of triceps brachii, extensor carpi radialis (ECR), and flexor carpi radialis (FCR). These muscles were chosen for the likely large contribution to tasks’ isometric contractions, as previously recommended to most accurately reconstruct synergies from a subset of muscles (Steele et al., 2013 (link)). Participants’ skin was prepared by rubbing a medical abrasive conductive paste (NuPrep, Weaver). Finally, electrode placement followed SENIAM and Cram’s recommendation guidelines (Hermens et al., 1999 ; Criswell, 2010 ). EMG signals were sampled at 2000 Hz, via a custom software interface.

EMG signals were recorded, during both experiments, from the following 13 upper limb muscles: Brachioradialis (BRD), Biceps brachii short head (BSH), Biceps brachii long head (BLH), Triceps brachii lateral head (TLT), Triceps brachii long head (TLN), Deltoid Anterior (DANT), Medial (DMED) and Posterior (DPOST) heads, Pectoralis Major (PM), Inferior head of the Trapezius (TRAP), Teres Major (TMAJ), and Latissimus Dorsi (LD). EMG signals were recorded through a Delsys Trigno system (Delsys, United States), sampled at 2000 Hz and synchronized with the load cell. EMG signals were first filtered in the 20 –400 Hz band by using a 3rd order digital Butterworth filter. The envelopes were then obtained by rectifying the signals and applying a low pass filter (3rd order Butterworth) with a cut-off frequency of 10 Hz. Before muscle synergies extraction, all the envelopes were amplitude normalized. The normalization was done with respect to the subject- and session-specific reference values calculated from the normalization block. During the normalization block, subjects reached three times to 8 targets spaced at 45°. The target associated with the maximal activation of each muscle was identified. The reference normalization value for each muscle was calculated as the average peak envelope value across the three repetitions of the target maximizing the muscle’s activity.