Temporal and spectral analyses of EEG microstate reveals neural effects of transcranial photobiomodulation on the resting brain

Abstract

Introduction: The quantification of electroencephalography (EEG) microstates is an effective method for analyzing synchronous neural firing and assessing the temporal dynamics of the resting state of the human brain. Transcranial photobiomodulation (tPBM) is a safe and effective modality to improve human cognition. However, it is unclear how prefrontal tPBM neuromodulates EEG microstates both temporally and spectrally.

Methods: 64-channel EEG was recorded from 45 healthy subjects in both 8-min active and sham tPBM sessions, using a 1064-nm laser applied to the right forehead of the subjects. After EEG data preprocessing, time-domain EEG microstate analysis was performed to obtain four microstate classes for both tPBM and sham sessions throughout the pre-, during-, and post-stimulation periods, followed by extraction of the respective microstate parameters. Moreover, frequency-domain analysis was performed by combining multivariate empirical mode decomposition with the Hilbert-Huang transform.

Results: Statistical analyses revealed that tPBM resulted in (1) a significant increase in the occurrence of microstates A and D and a significant decrease in the contribution of microstate C, (2) a substantial increase in the transition probabilities between microstates A and D, and (3) a substantial increase in the alpha power of microstate D.

Discussion: These findings confirm the neurophysiological effects of tPBM on EEG microstates of the resting brain, particularly in class D, which represents brain activation across the frontal and parietal regions. This study helps to better understand tPBM-induced dynamic alterations in EEG microstates that may be linked to the tPBM mechanism of action for the enhancement of human cognition.

Keywords: EEG microstate; Hilbert-Huang transform; electroencephalogram (EEG); empirical mode decomposition; transcranial photobiomodulation (tPBM).

Figure 1

(A) A photograph of our 1,064-nm laser used for the study. (B) A cartoon showing the EEG setup and the tPBM site on the participant’s right forehead. (C) Schematic diagram of the experimental protocol. A total of 45 subjects were randomly divided into two groups: active-sham or sham-active stimulation. Each experiment included EEG recordings of a 2-min baseline, an 8-min active or sham tPBM, and a 3-min recovery period. A minimum 1-week waiting period between two visits was required to avoid potential effects from active tPBM.

Figure 2

A flowchart for EEG microstate analysis. Panels (A–D) show steps of the time-domain EEG microstate analysis; Panels (F–H) show steps of the frequency-domain EEG microstate analysis. δ: delta band (0.5–4 Hz); θ: theta band (4–8 Hz); α: alpha band (8–13 Hz); β: beta band (13–30 Hz). Panels (E,I) represent the process for statistical analysis in the time and frequency domain, respectively.

Figure 3

EEG microstate topographies of 4 microstate classes. (A) Global microstate topographies obtained from both tPBM and sham sessions during four experimental periods (Pre, Stim1, Stim2, and Post). (B) Microstate topographies during the four temporal segments under Sham session. (C) Microstate topographies during the four temporal segments under active tPBM session.

Figure 4

Statistical comparisons of the microstate (A) occurrence (1/s) and (B) contribution (%) parameters among four temporal segments (Pre, Stim1, Stim2, and Post) for each of the four microstate classes, A, B, C, and D. These comparisons are made independently for the active and sham sessions. Statistical results were obtained by one-way repeated measures ANOVA and the post hoc pairwise comparisons with Tukey correction. Significant differences between a period pair are marked as “∗” for p < 0.05 after Tukey correction.

Figure 5

Transition probabilities between each pair of the four microstate classes and respective statistical comparisons under separate active and sham conditions. Statistical results were obtained by one-way repeated measures ANOVA and post hoc pairwise comparisons with Tukey’s correction. Significant differences between respective pairs are marked as “*” for p < 0.05, after Tukey’s correction.

Figure 6

Topographic maps of group-averaged (n = 45), baseline-normalized changes in microstate power in four microstate classes during tPBM/sham and post-tPBM/sham periods in the (A) delta band and (B) alpha band. Stars/crosses indicate clusters of electrodes with significant differences between the conditions (“*” for p_cluster < 0.01 and “x” for p_cluster < 0.05). The purple color of the stars/crosses in (A) indicates that the normalized delta powers in the tPBM session were significantly lower than those in the sham session. The red color of the crosses in (B) indicates that the normalized alpha powers of the tPBM session were significantly higher than those in the sham session.