Feasibility and reproducibility of electroencephalography-based corticokinematic coherence

Abstract

Corticokinematic coherence (CKC) is the phase coupling between limb kinematics and cortical neurophysiological signals reflecting cortical processing of proprioceptive afference, and is reproducible when estimated with magnetoencephalography (MEG). However, feasibility and reproducibility of CKC based on electroencephalography (EEG) is still unclear and is the primary object of the present report. Thirteen healthy right-handed volunteers (7 females, 21.7 ± 4.3 years) participated two separate EEG sessions 12.6±1.3 months apart. Participants' dominant and non-dominant index finger was continuously moved at 3 Hz for 4 min separately using a pneumatic-movement actuator. Coherence was computed between finger acceleration and three derivations of EEG signals: (1) average reference, (2) bipolar derivations, and (3) surface Laplacian. CKC strength was defined as the peak coherence value at the movement frequency. Intraclass-correlation coefficient values (0.74-0.93) indicated excellent inter-session reproducibility for CKC strength for all derivations and moved fingers. CKC strength obtained with EEG was ~2 times lower compared to MEG but the values were positively correlated across the participants. CKC strength was significantly (p<0.01) higher for bipolar (session-1 0.19±0.09; session-2 0.20±0.10) and surface Laplacian (session-1 0.22±0.09; session-2 0.21±0.09) derivations than for the average reference (session 1 0.10±0.04; session 2, 0.11±0.05). We demonstrated that CKC is feasible and reproducible tool to monitor proprioception using EEG recordings, although the strength of CKC was twice lower for EEG compared to MEG. Laplacian and bipolar (CP3-C1/CP3-C3 and CP4-C2/C4-FC2) EEG derivation(s) are recommended for future research and clinical use of CKC method.