Temporal Dynamics of Corticomuscular Coherence Reflects Alteration of the Central Mechanisms of Neural Motor Control in Post-Stroke Patients

Abstract

The neural control of muscular activity during a voluntary movement implies a continuous updating of a mix of afferent and efferent information. Corticomuscular coherence (CMC) is a powerful tool to explore the interactions between the motor cortex and the muscles involved in movement realization. The comparison of the temporal dynamics of CMC between healthy subjects and post-stroke patients could provide new insights into the question of how agonist and antagonist muscles are controlled related to motor performance during active voluntary movements. We recorded scalp electroencephalography activity, electromyography signals from agonist and antagonist muscles, and upper limb kinematics in eight healthy subjects and seventeen chronic post-stroke patients during twenty repeated voluntary elbow extensions and explored whether the modulation of the temporal dynamics of CMC could contribute to motor function impairment. Concomitantly with the alteration of elbow extension kinematics in post-stroke patients, dynamic CMC analysis showed a continuous CMC in both agonist and antagonist muscles during movement and highlighted that instantaneous CMC in antagonist muscles was higher for post-stroke patients compared to controls during the acceleration phase of elbow extension movement. In relation to motor control theories, our findings suggest that CMC could be involved in the online control of voluntary movement through the continuous integration of sensorimotor information. Moreover, specific alterations of CMC in antagonist muscles could reflect central command alterations of the selectivity in post-stroke patients.

Assessing Cognitive Workload in Motor Decision-Making through Functional Connectivity Analysis: Towards Early Detection and Monitoring of Neurodegenerative Diseases

“…Recently, coherence analysis using a new pre-processing computational technique [ 30 ] combining time–frequency analysis and the continuous wavelet transform has been introduced. This method enables processing data recorded during large joint or multi-joint movements [ 31 , 32 , 33 , 34 , 35 , 36 , 37 ]. There is still divergence in the results reported and the consensus regarding the specific frequency bands associated with motor control [ 38 ].…”

Exploring high-density corticomuscular networks after stroke to enable a hybrid Brain-Computer Interface for hand motor rehabilitation

“…Nevertheless, with respect to grasping patterns from healthy subjects, patients showed lower muscle specificity in all bands. This result is largely expected from a revision of CMC literature in stroke patients, showing involvement of proximal muscles to compensate for distal impairment [ 22 ] or higher contribution of antagonist muscles with respect to healthy subjects [ 14 , 23 ]. More generally, alterations of muscular involvement in post-stroke patients have largely been described through the phenomena of motor-overflow, co-activation of agonists and antagonists, spasticity and appearance of mirror movements [ 8 ].…”

“…However, the complexity of post-stroke recovery is such that several brain regions and muscles participate in post-lesional re-arrangements [ 17 – 20 ]. Lately, stroke-related CMC studies have broadened the observation to multi-channel recordings to describe complex phenomena such as the contralesional hemisphere contribution [ 21 ] or the abnormal recruitment of antagonists and proximal muscles [ 14 , 22 , 23 ]. All this evidence supports the potential role of CMC control feature in a rehabilitative BCI paradigm for its capability to encode both volitional control over movement and possible deviations from the physiological motor system activation, thus well beyond the purpose of increasing system classification performance.…”

Changes in intermuscular connectivity during active elbow extension reveal a functional simplification of motor control after stroke

“…Electromyographic continuous data were band-stop filtered at 49–51 Hz ( Krauth et al, 2019 ) to remove power line noise. As was done in Fauvet et al (2019 , 2021) , Glories et al (2021) , and Delcamp et al (2022) , the EMG signal was then 3-100 Hz band-pass filtered to keep the denoised part of the EMG signal energy that is necessary for reliable quantification of intermuscular coherence in the frequency band of interest in the present study. All the filters were fourth-order, zero-lag Butterworth types.…”