A flow chart showing five steps for (1) converting a time-domain series to a frequencydomain PSD, (2) repeating step 1 for all the channels, for each of the four rest/PVT epochs in two separate groups, (3) calculating channel-wise and channel-averaged (global) normalized PSD, (4) statistical testing if global normalized PSD values were significantly different between the rest and PVT epochs in each of TP1 and TP2, and (5) regrouping nPSD(f) to cover 5-min TP1 and TP2 for each channel and constructing sham-subtracted nPSD topography after multivariate statistical analysis for 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 nPSD values between S1 and S8 (right panel).

Contexts in source publication

Context 1

... normalized power spectral density during 10-min tPBM/sham epoch We temporally segmented the preprocessed data into resting and PVT phase as 2-minute and 3-minute windows, respectively. The power spectrum density (PSD) of each temporal segment was calculated using the native MATLAB function "pwelch" (sampling frequency of 512 Hz, discrete Fourier transform points of 4 s with 50% overlap) for each electrode and subject, as shown in Step 1 in Figure 2. Pwelch resulted in a 0.25-Hz spectral resolution, ranging from 0 Hz to 128 Hz. (1) converting a time-domain series to a frequencydomain PSD, (2) repeating step 1 for all the channels, for each of the four rest/PVT epochs in two separate groups, (3) calculating channel-wise and channel-averaged (global) normalized PSD, (4) statistical testing if global normalized PSD values were significantly different between the rest and PVT epochs in each of TP1 and TP2, and (5) regrouping nPSD(f) to cover 5-min TP1 and TP2 for each channel and constructing sham-subtracted nPSD topography after multivariate statistical analysis for TP1 and TP2 during the 10-min tPBM/sham epoch. ...

Context 2

... process produced a grouplevel normalized PSD (nPSD) at each EEG electrode in each frequency band. All nPSD values were also averaged across 64 EEG electrodes to obtain a global mean normalize PSD (nPSD ̅̅̅̅̅̅̅ ) in each epoch of rest and PVT in TP1 and TP2 for each participant in both the active and sham groups (Steps 2 and 3 in Figure 2). Furthermore, in TP1 and TP2, all the subjects underwent two resting and two PVT epochs. ...

Context 3

... t-tests were performed on paired nPSD(f) values between the 2-min rest and 3-min PVT epochs for each TP1 and TP2 using R studio [47] for both EEG data taken in S1 and S8, separately, and for each (sham and tPBM) group. This step was to examine whether we could pool the nPSD(f) values during the rest and PVT epochs together in each of TP1 and TP2 ( Step 4 in Figure 2). Significant differences in nPSD at the five frequency bands were set at a significance level of p<0.05. ...

Context 4

... 5 of Figure 2. In this case, cluster-based permutation testing examined whether the 4-week LEDtPBM could create significant alterations in EEG powers at the 5-f bands in certain brain regions when compared with those induced by the first 10-min tPBM in week 1. ...

Context 5

... this study, we observed that four weeks of LED-tPBM increased EEG nPSD globally in the theta, beta, and gamma bands ( Figure S2 in Supplementary Materials). It has been reported that attention and efficient processing of the task are often associated with theta [69,70], beta [70], and gamma bands [71][72][73][74]. ...