Learning at your brain's rhythm: individualized entrainment boosts learning for perceptual decisions

Abstract

Training is known to improve our ability to make decisions when interacting in complex environments. However, individuals vary in their ability to learn new tasks and acquire new skills in different settings. Here, we test whether this variability in learning ability relates to individual brain oscillatory states. We use a visual flicker paradigm to entrain individuals at their own brain rhythm (i.e. peak alpha frequency) as measured by resting-state electroencephalography (EEG). We demonstrate that this individual frequency-matched brain entrainment results in faster learning in a visual identification task (i.e. detecting targets embedded in background clutter) compared to entrainment that does not match an individual's alpha frequency. Further, we show that learning is specific to the phase relationship between the entraining flicker and the visual target stimulus. EEG during entrainment showed that individualized alpha entrainment boosts alpha power, induces phase alignment in the pre-stimulus period, and results in shorter latency of early visual evoked potentials, suggesting that brain entrainment facilitates early visual processing to support improved perceptual decisions. These findings suggest that individualized brain entrainment may boost perceptual learning by altering gain control mechanisms in the visual cortex, indicating a key role for individual neural oscillatory states in learning and brain plasticity.

Keywords: EEG; entrainment; learning; perceptual decisions; visual cortex.

Fig. 1

Experimental design and stimuli. A) Example stimuli comprising radial and concentric Glass patterns (stimuli are presented with inverted contrast for illustration purposes). Left: Prototype stimuli: 100% signal, spiral angle 0° for radial and 90° for concentric. Right: Stimuli used in the study: 25% signal, spiral angle 0° for radial and 90° for concentric. B) Trial design. Visual flicker (15 alpha cycles) was used to induce alpha entrainment. Each flash in the sequence was temporally separated by an interval equal to one cycle of each participant’s IAF. Following a blank interval at the end of the entrainment sequence (1–3 or 1.5–3.5 alpha cycles), the target stimulus was presented (200 ms). Participants were asked to judge whether the target stimulus was radial or concentric and indicated their decision with a button press. C) Experimental design. The entrainment frequency was either matched to the individual participant’s alpha frequency or was offset (nonMatched) by ±1 Hz. The onset of the target stimulus was set either at the peak or trough of the oscillation induced by the visual flicker by manipulating the interval after the entrainment sequence: for 10 Hz stimulation at the peak, the interval was 100, 200, or 300 ms; for 10 Hz stimulation at the trough, the interval was 150, 250, or 350 ms. These values were scaled according to the participant’s IAF. The solid line indicates the hypothesized trajectory of the entrained alpha oscillation during the visual flicker sequence. The dashed line reflects the hypothesized continuation of the entrained alpha oscillation after the flicker sequence has ended, with stimuli shown at all possible presentation times.

Fig. 2

Behavioral performance. A) Performance (% correct) across blocks in Session 1. Open circles indicate mean accuracy (percentage of correct responses) per block of trials (~100 trials per point), for each of the 3 intervention groups (T-Match, T-misMatch, and P-Match). Accuracy data across blocks were fitted with a logarithmic function (solid lines) to estimate learning rate. B) Mean learning rate (i.e. slope of logarithmic fit) across participants per group. Bars show mean learning rate per group, as estimated from fitting the individual accuracy data. Error bars show ±1 SEM.

Fig. 3

Entrainment effects on alpha power. A) Time series showing amplitude of alpha envelope in the on-target frequencies within the trial epoch. Gray highlighted regions show the 3 time windows of interest for further analysis: entrainment window, pre-stimulus, and post-stimulus (from L-R). Solid lines show the mean per group, shaded regions show ±1 SEM. B) Barplots showing the comparison of mean alpha amplitude within the entrainment window, for on-target (solid color) and off-target (hatched) alpha frequencies across groups. Error bars show ±1 SEM. C) Mean alpha amplitude across participants within the entrainment window for on-target frequencies is shown across blocks for each experimental group. Error bars show ±1 SEM.

Fig. 4

Entrainment effects on alpha phase. Circular mean phase angle for stimulus onset across participants (solid lines) per group; individual values are shown in the corresponding circular histogram. Data are shown for on- versus off-target frequency for 2 comparisons: Trough-Match versus Peak-Match (left column) and Trough-Match versus Trough-nonMatch (right column).

Fig. 5

Entrainment effects on visual evoked potentials. A) Posterior evoked potential for all trials, time-locked to the onset of the target stimulus. Top panel shows time series for central occipital electrodes and the gray shaded region indicate the N1 search region. Bottom panel shows the time series for the lateral occipital electrodes and gray shaded areas indicate P1 and P2 search windows. Data are shown for each group; solid line indicates mean amplitude; shaded regions indicate ±1 SEM. B) Topographies for the amplitude of the evoked response averaged within the search windows for each component (P1, N1, and P2 respectively). Black dots indicate electrode sites. C) Left panel: Barplots show mean latency of evoked potential response across participants, following target stimulus onset, for each entrainment group. Top right panel shows the between-group comparison of latency values for the N1 component. Bottom panel shows the mean component amplitude across all components (N1, P1, P2). Error bars indicate SEM across participants.