Repeated Transcranial Photobiomodulation with Light-Emitting Diodes Improves Psychomotor Vigilance and EEG Networks of the Human Brain

Abstract

Transcranial photobiomodulation (tPBM) has been suggested as a non-invasive neuromodulation tool. The repetitive administration of light-emitting diode (LED)-based tPBM for several weeks significantly improves human cognition. To understand the electrophysiological effects of LED-tPBM on the human brain, we investigated alterations by repeated tPBM in vigilance performance and brain networks using electroencephalography (EEG) in healthy participants. Active and sham LED-based tPBM were administered to the right forehead of young participants twice a week for four weeks. The participants performed a psychomotor vigilance task (PVT) during each tPBM/sham experiment. A 64-electrode EEG system recorded electrophysiological signals from each participant during the first and last visits in a 4-week study. Topographical maps of the EEG power enhanced by tPBM were statistically compared for the repeated tPBM effect. A new data processing framework combining the group's singular value decomposition (gSVD) with eLORETA was implemented to identify EEG brain networks. The reaction time of the PVT in the tPBM-treated group was significantly improved over four weeks compared to that in the sham group. We observed acute increases in EEG delta and alpha powers during a 10 min LED-tPBM while the participants performed the PVT task. We also found that the theta, beta, and gamma EEG powers significantly increased overall after four weeks of LED-tPBM. Combining gSVD with eLORETA enabled us to identify EEG brain networks and the corresponding network power changes by repeated 4-week tPBM. This study clearly demonstrated that a 4-week prefrontal LED-tPBM can neuromodulate several key EEG networks, implying a possible causal effect between modulated brain networks and improved psychomotor vigilance outcomes.

Keywords: EEG; PVT; electroencephalography; light-emitting diodes; psychomotor vigilance task; repeated tPBM; tPBM; transcranial photobiomodulation.

Figure 1

A photograph of (a) the LED-tPBM unit used for the study, and (b) a participant wearing an EEG cap and performing PVT while the LED light was delivered to (c) the right forehead. (d) Diagram of the experimental protocol showing a total of eight sessions of the experiment, S1 to S8, over four weeks. EEG measurements were taken concurrently while the participants performed PVT only in S1 of Week 1 and S8 of Week 4. (e) A schematic illustration for each of the eight sessions. This between-group experimental design included 11 subjects randomly assigned to either the active or the sham group. The entire experiment lasted for 20 min. During the first ten minutes, the participants in both groups were not subjected to any treatment. During the last ten minutes, the active (n = 11) or sham (n = 11) group was administered with true tPBM or sham treatment, respectively, on the right forehead. One 2 min resting epoch and 3 min PVT epoch together consisted of one time period (TP). Participants in each group underwent TP1 and TP2 at each visit. PVT was performed twice during the 10 min tPBM period per visit. The respective reaction times were recorded twice per visit, resulting in a total of 16 reaction times measured over 4 weeks (2/visit × 2 visit/week × 4 weeks = 16).

Figure 2

A flow chart showing the five steps for (1) converting a time-domain series to a frequency-domain PSD, namely, PSD(f), where f covers five EEG frequency bands, (2) repeating Step 1 for all the channels for each of the four rest/PVT epochs in the two separate groups, (3) calculating the channel-wise and channel-averaged (global) normalized PSD(f), (nPSD(f) and nPSD(f)₆₄), (4) statistical testing on whether the global nPSD(f) values were significantly different between the rest and PVT epochs in each of TP1 and TP2, and (5) regrouping nPSD to cover the 5 min TP1 and TP2 for each channel in each of five frequency (5-f) bands, as well as constructing sham-subtracted nPSD topography for each of 5-f bands after multivariate statistical analysis for the TP1 and TP2 during the 10 min tPBM/sham epoch. This analysis was performed separately to examine the acute effect during Session S1 (left panel), and to determine the repeated effect of the 4-week LED-tPBM by comparing the nPSD values between S1 and S8 (right panel).

Figure 3

An EEG data processing flowchart showing the eight steps through which to obtain 2D- and 3D-brain EEG networks, as well as the respective alterations induced by 4-week LED-tPBM. All steps were performed using MATLAB except Step 6, which was performed using eLORETA software. See text and Section SA of the Supplementary Materials for details on each processing step.

Figure 4

Effect of repeated 10 min LED-tPBM on the reaction time for the sham (n = 11; blue) and tPBM (n = 11; red) groups. Both dotted lines are linear fits to the sham (blue) and tPBM (red) groups, with p-values of 0.07 and 0.002, respectively. Each data point reflects the mean reaction time averaged over each respective group during the 3 min PVT in each session from S1 to S8 over four weeks. Note that, in each session (e.g., S1), two reaction times were recorded in TP1 and TP2, leading to a total of 16 data points over 4 weeks.

Figure 5

Normalized EEG power spectrum density plots averaged over all 64 channels (i.e., nPSD₆₄) from the sham (blue curve) and tPBM (red curve) groups during (a) TP1 and (b) TP2. Normalization was calculated with respect to the 2 min baseline taken at the beginning of the experiment without light stimulation (see Figure 1e). The x-axis denotes the EEG frequency in Hz, and the y-axis represents the percentage change in nPSD₆₄ relative to the PSD₆₄ during the 2 min baseline. (c) Topographic maps of the nPSD at each of 5-f bands. This also shows the statistical comparisons of the channel-interpolated topographical alterations in the nPSD between the sham (n = 9) and tPBM (n = 7) groups during TP1 and TP2 in Week 1. The columns indicate the delta, theta, alpha, beta, and gamma bands. The color bar indicates the percentage change in the nPSD with respect to the baseline for each electrode channel. The “*” marks a p value < 0.05 for statistical significance of nPSD between tPBM and sham groups after multivariate comparison correction.

Figure 6

(a) Ranks and weights of all the 64 SVD components after gSVD. The x-axis indicates the ranks of the components, while the y-axis denotes the weight of each component. The 12 most-weighted components are marked by red dots and selected for further data processing. The total weight of the 12 most-weighted components (red dots) takes 70% of the total weight (all the dots). (b) The group-averaged PSD curves from the sham (blue) and tPBM (red) groups during TP2 for component two (SVD #2). Blue vertical lines mark the five EEG frequency bands; namely, delta (δ: 0.5–4 Hz), theta (θ: 4–7 Hz), alpha (α: 7–13 Hz), beta (β: 13–30 Hz), and gamma (γ: 30–70 Hz). (c) A correlation matrix to show the group-averaged PCC values for every pair of the 12 networks. The vertical and horizontal axis depict the 12 gSVD components, and the color bar represents the PCC between each pair of networks. Since all the self-correlations were meaningless, all the values along the diagonal line were marked as N/A (i.e., not applicable).

Figure 7

The 2D rEP maps of the 12 identified gSVD components (labeled by SVD#) with 3D source localizations of cortical current density. For each of the two panels, the second left most column shows the 2D rEP map; the middle three columns reveal the axial, sagittal, and coronial views for each rEP map, with a unit of cm in all three dimensions; and the two right most columns depict the top and side views of the left and right hemispheres for the 3D rendered brain templates. The yellow color in each brain model indicates the binarized cortical current density for a threshold of >75% of the maximum neuronal activity in the respective brain models. The color bars show the rEP of the ‘dipole’ across the scalp (no unit).

Figure 8

Group-level EEG power changes, Δnp, between S8 (in Week 4) and S1 (in Week 1) for the 12 brain networks in (a) delta, (b) theta, (c) alpha, (d) beta, and (e) gamma bands during TP1 from the sham (blue) and tPBM (red) groups. The standard error of the mean is represented by the error bars. The significant differences of Δnp between the two groups were determined using the two-sample non-parametric test at the significance level of p < 0.05 and are indicated by ‘*’.

Figure 9

Group-level EEG power changes, Δnp, between S8 (in Week 4) and S1 (in Week 1) for the 12 brain networks in the (a) delta, (b) theta, (c) alpha, (d) beta, and (e) gamma bands during TP2 from the sham (blue) and tPBM (red) groups. The standard error of the mean is represented by the error bars. The significant differences of Δnp between the two groups were determined using the two-sample non-parametric test at the significance level of p < 0.05 and are indicated by ‘*’.