Nonlinear Coupling between Cortical Oscillations and Muscle Activity during Isotonic Wrist Flexion

Abstract

Coupling between cortical oscillations and muscle activity facilitates neuronal communication during motor control. The linear part of this coupling, known as corticomuscular coherence, has received substantial attention, even though neuronal communication underlying motor control has been demonstrated to be highly nonlinear. A full assessment of corticomuscular coupling, including the nonlinear part, is essential to understand the neuronal communication within the sensorimotor system. In this study, we applied the recently developed n:m coherence method to assess nonlinear corticomuscular coupling during isotonic wrist flexion. The n:m coherence is a generalized metric for quantifying nonlinear cross-frequency coupling as well as linear iso-frequency coupling. By using independent component analysis (ICA) and equivalent current dipole source localization, we identify four sensorimotor related brain areas based on the locations of the dipoles, i.e., the contralateral primary sensorimotor areas, supplementary motor area (SMA), prefrontal area (PFA) and posterior parietal cortex (PPC). For all these areas, linear coupling between electroencephalogram (EEG) and electromyogram (EMG) is present with peaks in the beta band (15–35 Hz), while nonlinear coupling is detected with both integer (1:2, 1:3, 1:4) and non-integer (2:3) harmonics. Significant differences between brain areas is shown in linear coupling with stronger coherence for the primary sensorimotor areas and motor association cortices (SMA, PFA) compared to the sensory association area (PPC); but not for the nonlinear coupling. Moreover, the detected nonlinear coupling is similar to previously reported nonlinear coupling of cortical activity to somatosensory stimuli. We suggest that the descending motor pathways mainly contribute to linear corticomuscular coupling, while nonlinear coupling likely originates from sensory feedback.

Unveiling neural coupling within the sensorimotor system: directionality and nonlinearity

“…Recently, we proposed a frequency domain approach namely cross‐spectral coherence (CSC) to investigate the nonlinear corticomuscular interaction (Yang et al ., ). CSC is a generalised coherence framework for quantifying nonlinear coupling between signals across different frequency bands as well as the linear coupling within the same frequency band (Yang et al ., ).…”

“…Using CSC and independent component analysis, we assessed both linear and nonlinear interaction between muscle activity and multiple brain sources in healthy participants during an isotonic wrist flexion task (Yang et al ., ). In consistent with previous studies, we found beta‐band peak in the linear corticomuscular interaction for both motor and sensory‐related cortices, that is primary sensorimotor areas (S1‐M1), premotor area (PMA), supplementary motor area (SMA) and posterior parietal cortex (PPC).…”

Neuromechanical Modelling of Articulatory Movements from Surface Electromyography and Speech Formants

“…There are several important considerations in its current formulation (H4: relationship between motor unit actions and force). [36][37][38][39][40][41][42] The question of why force is related to the rectified sEMG, and not to sEMG, apparently has to see with the way in which sEMG is recorded, using pairs of electrodes symmetrically placed at both sides of the neuromotor innervation zone on the muscle, as suggested by the experts. 37,39 The force exerted by a muscle is a consequence of the neuromotor stimulation of the muscular fibers by impulse-like action potentials, 43 which accumulate on the muscular cell membranes as a de facto integration of the positive peaks of individual actions, therefore the muscular force appearing responds to the rectified integral of sEMG recordings (r m (t)).…”

Impaired corticomuscular coherence during isometric elbow flexion contractions in humans with cervical spinal cord injury

“…1D) was then obtained with parameters 'nvoice' (scale resolution of the wavelet), 'J1' (number of scales) and 'wavenumber' (Morlet mother wavelet parameter) set, respectively, to 0.125, 871 and 7 to yield accurate identification of oscillatory activity from 0.13 Hz to 114.39 Hz in 1.07 Hz step. EMG signals were not rectified to properly model EMG time series as centered Gaussian processes (Bigot et al, 2011;Charissou et al, 2016) and not to lose important information contained in the EMG signal spectrum (Neto & Christou, 2010;McClelland et al, 2012;Yang et al, 2016). The wavelet cross-spectrum ( Fig.…”

“…where S EEG, EEG (x, u) is the wavelet cross-spectrum between EMG and EEG signals (Eqn 2), and S EMG (x, u) and S EEG (x, u) are the wavelet auto-spectrum of each EMG and EEG signal (Eqns 3 and 4). The magnitude of the CMC was calculated from 0 to 1 bounded corticomuscular values as the volume under the time-frequency map of magnitude-squared coherence only where the wavelet crossspectrum between the EEG and EMG signals was detected as significant (Charissou et al, 2016;Yoshida et al, 2017). Over the [+3 + 6] s period of interest, the magnitude of the CMC was computed in two frequency bands of interest: [8-13] Hz (CMC 8-13 ) (Christou et al, 2007) and [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31] Hz (CMC 13-31 ) .…”