2.1.

Participants

A total of 20 healthy human participants with an average ($\pm \mathrm{SD}$) age of $26.8\pm 8.8$ years were recruited from the local community of the University of Texas at Arlington, Texas. Specifically, 7 females ($24.0\pm 4.1$ years of age) and 13 males ($28.2\pm 10.3$ years of age) participated in the study; the two gender groups had no significant age difference (with a two-tailed $t$-test, $p>0.2$). The exclusion criteria included: (1) previous diagnosis with a psychiatric disorder, (2) history of neurological disease, (3) history of severe brain injury, (4) history of violent behavior, (5) prior institutionalization/imprisonment, (6) current intake of any psychotropic medicine or drug, (7) previous diagnosis with diabetes, (8) history of smoking, (9) excessive alcohol consumption, or (10) pregnancy. Exclusion of diabetic patients was required by the manufacturer of the laser (Cell Gen Therapeutics LLC, Dallas, Texas); lack of prior psychiatric diagnosis was ascertained by self-report. The study protocol was approved by the Institutional Review Board (IRB) at the University of Texas in Arlington. Informed consent was obtained prior to all experiments.

2.2.

Experimental Setup for Simultaneous EEG-tPBM Measurements

We employed an FDA-cleared, 1064-nm, continuous-wave (CW) laser (Model CG-5000 Laser, Cell Gen Therapeutics LLC, Dallas, Texas) used in previous studies.^13–18,30 Tables 1Table 2–3 list the tPBM parameters used in this study following the recommended format given in Ref. 31. The laser’s aperture delivered a well-collimated beam with an area of $13.6{\mathrm{cm}}^2$. To avoid thermal sensation for a more truthful sham-controlled study, we decreased the laser power to 2.2 W from 3.4 W used in previous studies.^13–18,30 Indeed, all participants reported no thermal sensation. This laser power of 2.2 W resulted in a power density of $0.162\mathrm{W}/{\mathrm{cm}}^2$, energy density of $\sim 107\mathrm{J}/{\mathrm{cm}}^2$, and a total energy dose of 1452 J over the 11-min tPBM duration ($2.2\mathrm{W}\times 660\mathrm{s}=1452\mathrm{J}$). Note that the optical energy density (or fluence) received on the cortical region would be only 1% to 2% of that delivered on the scalp by the laser aperture, based on recent literature.^7,32–34 For sham stimulation, the laser power output was set to 0.1 W, and a black cap was also placed in front of the aperture to further block the laser. The corresponding power density under the sham stimulation was further tested to be $0\mathrm{mW}/{\mathrm{cm}}^2$ by a sensitive power meter (Model 843-R, Newport Corporation, Massachusetts) to ensure complete blockage of laser light.

Table 1

Device information.

Table 2

Irradiation parameters.

Table 3

Treatment parameters.

The experimental setup of the simultaneous EEG-tPBM measurements is shown in Figs. 1(a) and 1(b). The EEG data were collected using a 64-channel EEG instrument (ActiveTwo, Biosemi, the Netherlands). Channel 48, Cz, was used as the reference. The tPBM was applied by noncontact delivery to the right forehead of each subject at a frontal site (near the Fp2 location of the international 10–10 system); the same laser light delivery location demonstrated to enhance cognitive performance and local oxygen metabolism in our previous studies.^14–16,18 Laser protection goggles with extra black-tape rims were worn by the subject and the experimental operator throughout the entire experiment, before, during, and after tPBM. To prevent sleepiness or drowsiness, the subjects kept their eyes open during the EEG-tPBM measurements after the laser aperture was well positioned on their forehead. Each subject wore an EEG cap with 64 electrodes positioned according to the standard 10–10 EEG electrode placement.³⁵

Fig. 1

(a) A highly schematic drawing of the setup, (b) a photo of actual setup of the EEG-tPBM experimental setup, and (c) illustration of the study protocol for EEG-tPBM and EEG-sham experiments.

2.3.

Experiments of Simultaneous EEG-tPBM

Figure 1(c) shows the experimental design for both sham and active tPBMs. Experiments consisted of a 2-min baseline period (Bl), an 11-min laser stimulation period (T1–T11), and a 3-min recovery period (R1–R3). All 64-channel EEG data were acquired throughout the 16-min duration. Each subject received both active and sham stimulations on the same day, with sham taken first, followed by active stimulation. All subjects were unaware that one of the sessions consisted of sham stimulation. After each experiment, each subject was asked about perceptions of heat from the stimulation.

2.4.

EEG Data Analysis

We conducted data analysis on 64-channel EEG time series in several steps.

2.4.4.

Time-resolved, baseline-normalized topography

For each frequency band, we created time-dependent, 64-channel (i.e., $64\times 1$) power density vectors, $P_{\mathrm{tPBM}}$ or $P_{\mathrm{sham}}$, for either tPBM or sham experiment during the stimulation (11 min) and recovery (3 min) period, as well as a corresponding baseline power density vector, $P_{\text{base}}$ (based on the second-minute baseline data). Baseline-normalized, tPBM-induced (or sham-induced) increases in power density were quantified as $\mathrm{n}{P}_{\mathrm{tPBM}}=P_{\mathrm{tPBM}}/P_{\text{base}}-1$ (or $\mathrm{n}{P}_{\mathrm{sham}}=P_{\mathrm{sham}}/P_{\text{base}}-1$) for each of the 14 time-dependent periods during and after tPBM (i.e., T1–T11 and R1–R3). This data processing routine is shown in Fig. 2. Next, after close inspection on the magnitudes of time-resolved, baseline-normalized EEG power density for each EEG channel at each frequency band, we excluded three subjects as outliers. The exclusion criterion was that any subject whose laser-induced power alteration (i.e., increase or decrease) in any electrode was equal to or larger than four standard deviations away from the sample/group mean would be excluded as an outlier, regardless of any channel or/and any frequency band. In other words, we calculated the sample/group mean (${\mu }_{\text{sample}}$) and standard deviation (${\sigma }_{\text{sample}}$) of the power ratios, for each channel at each frequency, for all the subjects. The ratios that were larger than ${\mu }_{\text{sample}}+4{\sigma }_{\text{sample}}$ or smaller than ${\mu }_{\text{sample}}-4{\sigma }_{\text{sample}}$, were regarded as outliers, and the corresponding participant was removed from further data processing.

Fig. 2

A flow chart for EEG data analysis to obtain topographic maps of EEG power spectral density differences between active and sham tPBMs, followed by time-dependent the t-test for statistical comparison. This process routine was used for each of the five frequency bands.

4.1.

tPBM Modulates Alpha and Beta Power

Figure 4 illustrates that the 11-min tPBM delivered to the right forehead of 20 human subjects (three outliers excluded) results in up to 20% spatially broad increases of human alpha and beta band powers. Figure 4(a) reveals that, in the alpha band, the scalp EEG oscillation powers increased progressively after the laser stimulation was initiated relative to the sham condition. Furthermore, the strongest increases in alpha oscillation power appeared during 8 to 11 min with a dominant front-to-back pattern bilaterally. Figure 4(b) shows a similar trend of rhythm power increases for the beta band, particularly still remaining the dominant front-to-back pattern during the later period of tPBM (i.e., during 8 to 11 min).

Together the figures reveal several key features: (1) laser-induced increases of EEG power density were frequency-dependent since the statistically significant increases occurred only at the alpha and beta bands, with very strong effects ($\text{maximum effect size}\cong 1.5$); (2) laser-induced alterations were cumulatively time-dependent since the changes took place on a time scale of minutes for both alpha and beta bands, and they were largest during 8 to 11 min after laser onset; and (3) the alterations were location-dependent, based on the observations that significant increases in power density appeared to be front-to-back on the scalp 3 min after stimulation onset for the alpha and beta bands, with increases remaining front-to-back for the alpha band and for the beta band after 7 min of the stimulation onset.

It is commonly regarded that an increase in alpha power alone implies a less attentive state and less arousal. It might be argued that the increase of alpha power seen in Fig. 4(a) could result from the subjects’ sleepiness as they sat in the experiment for a while. If subjects get sleepy, alpha increases but not both alpha and beta as produced by the laser as compared to the sham with the same temporal period. Therefore, the tPBM effect is not the same as simply lowering arousal level. This difference may be important for the cognitive benefit produced by the tPBM. It is also reported that alpha and beta rhythms of brain oscillations result from synchronous electrical activity of thalamocortical neurons, with alpha more characteristic of quiet awake states and beta of alert states.^42,43 More specifically, alpha waves are robustly modulated during cognitive processes and play an important role in the integration and communication among different brain rhythms during brain activity.⁴⁴ Beta oscillations are also relevant to cognitive processing but to a lesser degree than alpha waves.^45,46 Our results reveal that tPBM is able to create large effects on not only alpha but also beta oscillations. All of these observations lead us to speculate that the laser-induced large effect in both alpha and beta powers may be a mechanistic link to tPBM-induced improvement and enhancement of human cognition that have been repeatedly reported in sham-controlled studies by Gonzalez-Lima et al. and others in recent years.^13–16,30 As supporting evidence for this speculation, many studies in neurofeedback research have observed improvement of cognitive performance while the upper alpha amplitude increased⁴⁷ by neurofeedback training.^47–51

We have also noticed that prefrontal tPBM has created a dominant front-to-back pattern in both alpha and beta rhythm powers, with a few minutes delay in the beta power increase. Since there is no prior literature on such observation, one possible speculation is to attribute this pattern to tPBM-induced enhancement in large-scale brain circuits in one or more prefrontal-driven brain networks, such as the frontoparietal central executive network and the default mode network (DMN) (see Sec. 4.2 next for more details). The former network is known for initiation and modulation of cognitive control.⁵² If prefrontal tPBM is proven to be able to modulate this network, it would reveal an important electrophysiological mechanism of action of tPBM at the brain circuit level and thus lead to a better understanding of tPBM-based interventions for both neuroscience research and clinical applications. Our speculation needs to be investigated and tested in future studies.

4.2.

tPBM Holds Potential for Photobiomodulation of the Human Default Mode Network

The DMN is a network of brain areas activated during the resting state; the DMN becomes suppressed once certain tasks are involved.⁵³ Malfunction and disconnection of the DMN have been related to the cognitive dysfunctions seen in many psychiatric and neurological conditions, such as depression, anxiety, and Alzheimer’s disease.^53–55 The abnormal connectivity of the DMN in those conditions makes the brain unable to stay in a normal resting state.⁵³ Increases of EEG power density in the anterior frontal regions indicate that tPBM may have the potential or ability to strengthen and stimulate neural functions of the anterior DMN. In particular, the T-maps of EEG alpha power during 8 to 11 min [Fig. 4(a)] illustrate that the scalp area covering (entirely or partially) the anterior DMN is significantly modulated by tPBM. We note, however, that brain imaging modalities with higher spatial resolution than that of EEG will be needed to pinpoint the areas modulated by tPBM. The modulatory potential of tPBM on the DMN suggests a promising prospect for tPBM to promote DMN functions in healthy people and perhaps reverse DMN changes found in various psychiatric and neurological conditions such as depression, anxiety, and Alzheimer’s disease.

4.3.

Continuous-Wave versus Pulsed tPBM

The stimulation light used in our current^17,18 and previous^13–15 studies was CW laser, and most light sources employed in other tPBM investigations were also nonpulsed lasers or LEDs. To date, it is still unclear whether CW PBM is better or worse than pulsed PBM for different uses (e.g., physical therapy, chiropractic, pain reduction, sports medicine),⁵⁶ and very limited reports with pulsed tPBM can be found. For example, a human study with a 10-Hz pulse frequency was recently published,⁵⁷ and several preclinical studies explored treatment outcomes of pulsed tPBM on different types of neurological conditions with different pulse frequencies.^58–62 Table 4 lists corresponding pulsed frequencies, studied species, and related references. In principle, one advantage of CW tPBM is to deliver more optical energy/fluence, and thus more dosage in a given time, but one disadvantage is to possibly create more heat on the skin surface. The answer to the question of whether CW tPBM is better or worse with respect to pulsed tPBM for affecting EEG rhythms or other applications remains to be further investigated for the entire tPBM field.

Table 4

Studies using pulsed tPBM.

4.4.

Limitations of the Study and Future Work

Since this was the first sham-controlled, time-resolved topographic study of EEG-tPBM in healthy humans, it had a few limitations that could be overcome in future investigations. First, to prevent the subjects from falling asleep during the EEG measurements, the subjects kept their eyes open during the measurements while wearing safety goggles with a layer of light-absorbing sheet over the edges for eye protection. However, in order to investigate tPBM-induced effects thoroughly on alpha oscillations, it might be appropriate to conduct future experiments with the eyes closed, enabling detection of stronger alpha activity. Second, while the EEG data in this study were acquired using a conventional 10–10 system of scalp electrode placement, the EEG cap was positioned more posteriorly than the regular setup in order to have an adequate space for a 4-cm-diameter laser stimulation site [see Fig. 1(b)]. The shift of the cap was $\sim 2\mathrm{cm}$ up along the anterior–posterior midline (or the nasion to inion line), which would cause inconsistency between the conventional 10–10 system electrode sites versus the actual ones used in this study. This mismatch would prevent us from identifying individual scalp locations/regions that have significant EEG power responses to tPBM with high accuracy. A solution to this limitation may be to coregister the actual EEG electrodes with the conventional electrode placement using a three-dimensional digitizer to measure several key anatomical landmarks, followed by coordinate transformation. Third, we speculated that the tPBM-induced increase in both alpha and beta power density may be a mechanistic link to tPBM-induced improvement and enhancement of human cognition that have been reported in sham-controlled studies in recent years.^13–16,30 This speculation needs to be further investigated and confirmed by combining EEG-tPBM measurements with tPBM-enhanced human cognition tests in the future. Last, we cannot rule out that heating did not play a role in the electrophysiological changes observed here. Cortical temperature is known to change with neural activation by $\sim 0.2\text{-}\mathrm{deg}$ Celsius.⁶³ It is possible that the reverse effect—exogenous heat alters neural activity—could have contributed to the observed effects. On the other hand, we have previously shown that thermal stimulation of the forehead does not lead to the increased cortical CCO or hemodynamic changes with tPBM.⁶⁴ In the future, a similar thermal stimulation study of EEG and computational simulations may be possible approaches to quantify or estimate tPBM-induced versus thermal effects on the EEG and thus be able to disambiguate the effects of photobiomodulation and tissue heating.