Article Figures & Data
Figures
- Figure 1.
Effect of neurofeedback training on alpha power in a representative participant. A, Details of the experimental paradigm. Three types of trials, namely valid (red), invalid (green), and constant (blue) were presented for five sessions, each consisting of 12 trials. Each trial was 50 s long. Each block started with a calibration stage (data not shown). The first session, in which invalid trials were not presented, was not used for analysis. B, Time-frequency spectogram of a single valid trial showing change in power from baseline (computed during the calibration stage). Broken lines at 8 and 13 Hz indicate the alpha range. C, Change in instantaneous alpha power for the same trial as in B (brown trace; left y-axis). Dotted orange line depicts alpha power smoothed by averaging across the previous 5 s, which was used to set the frequency of the feedback tone (green trace; right y-axis). D, Raw alpha power versus trial number during calibration (thick black line; same value for each block of 12 trials), eyes open (open circles), and eyes closed state (solid triangles). Regression lines between raw alpha power versus trial number (13–60) are shown for eyes open (gray trace) and eyes closed states (brown trace). Corresponding slopes and p values are indicated in the panel in respective colors. Error bar indicates SEM. E, Change in alpha power with respect to time for three types of trials: valid (red), invalid (green), and constant (blue), averaged over trials 13–60 (24 valid, 12 invalid, and 12 constant trials). Regression lines plotted between mean change in alpha power and time (21–50 s) are also shown in corresponding colors. Average change in power in decibels (between 21 and 50 s) ±SEM for the three types trials are indicated at top right corner (in corresponding colors). Slopes of the regression lines along with their p values are indicated at bottom right corner.
- Figure 2.
Modulation of alpha power for three types of trials for all participants. Same as Figure 1E, separately for each of the 24 participants, in descending order of significance of the difference between change in alpha power between valid and invalid trials (estimated using t test and shown in the plots, along with the mean changes in alpha power for the three trial types).
- Figure 3.
Effect of neurofeedback depends on sustenance of alpha power. Scatterplot shows difference in power between valid and invalid trials ( Δ power) and slope for the constant trials for all the 24 participants. Thick black line shows a regression line; slope and p values are indicated in the panel. Participants 1–6, for which mean change in alpha power between valid and invalid trials was significant after Bonferroni correction, are indicated using filled black triangles. Participants 7–11, for which the difference was significant before Bonferroni correction, are indicated using gray triangles. Remaining participants are indicated using open circles. Error bar indicates SEM.
- Figure 4.
Subjective experience of the neurofeedback training for all participants. Scatterplot showing difference in alpha power between valid and invalid trials with respect to (A) feedback scores (FSs), (B) relaxation score (RS), and (C) disturbance scores (DSs). For each plot, regression line is shown in thick black, and slope and p values are indicated in the panel. Same markers as in Figure 3. Error bar indicates SEM.
In this issue
