Example trial sequence and rhythmic visual stimulation.

(A) In each trial, infants saw a cartoon monster inside a white circle for 2 s (ISI: 0.7 to 0.9s). (B) Stimuli were presented periodically at discrete eight frequencies (2, 4, 6, 8, 10, 15, 20, 30Hz) and in broadband sequences.

Visual system response to periodic stimulation at 2, 4, 6, 8, 10, 15, 20, 30Hz.

(A) The neural response spectra as a function of stimulation frequencies at parietooccipital electrodes. The signal-to-noise ratio (SNR) was above chance (SNR = 1) for the first, second and third harmonic responses, and the 4Hz rhythm for most stimulation frequencies (see Table 1 for statistics). (B) Topographic maps display averaged responses at first, second and third harmonic, and at 4Hz across all stimulation frequencies.

Visual system response to periodic stimulation at 2, 4, 6, 8, 10, 15, 20, 30Hz.

Mean SNRs and standard deviations are presented for the first, second and third harmonics harmonic, and the ∼4Hz response. For the 4Hz response, the maximum frequency (i.e., 4 or 4.5Hz) is reported. Statistics are based on one-sample t-tests against baseline (SNR = 1). * significant at p < .05, ** significant at p < .01, ***significant at p < .001.

Visual system response to broadband visual stimulation.

(A) The IRF was estimated by cross-correlating the broadband sequences and the EEG signal of the same trials. The surrogate was calculated as the cross-correlation with random trials (blue). The IRF, surrogate and corresponding spectral representations are shown for two representative participants (P). (B) IRF and surrogate together with their spectral representation across all infants. (C) The spectral difference between the IRF and the surrogate and (D) the topography of the IRF at 4Hz.

Response to broadband visual stimulation after removing the 3.5 to 4.5Hz component of the broadband sequence.

(A) IRF of the band-stop filtered stimulus sequence (purple) compared to the IRF of the full-spectrum stimulus (red). (B) The corresponding TFRs show reduced 4 Hz power after filtering. (C) Power spectra (2–10 Hz) of the 0 to 1 s time window for the band-stop filtered and full spectrum IRFs.

Visual system response to periodic stimulation at 2, 4, 6, 8, 10, 15, 20, 30Hz and to broadband stimulation in N = 7 adults.

(A) The neural response spectra as a function of stimulation frequencies at parietooccipital electrodes. The SNR was above chance (SNR = 1) for first, second and third harmonic responses, and for the individual peak alpha frequency for most stimulation frequencies. (B) Topographic maps display averaged responses at first, second and third harmonic, and at the individual peak alpha frequency across all stimulation frequencies. (C) Response to broadband stimulation for two representative participants -IRF (red) and surrogate (blue) (D). TFR of IRF (top) and the surrogate function (bottom) across all adult subjects and (E) their difference.

SNR response to periodic visual stimulation at 2, 4, 6, 8, 10, 15, 20, 30Hz at frontal electrodes.

The 4Hz response was particularly pronounced for frontal sites (Fp1, Fz, F3, F7, F4, F8, Fp2, AF7, AF3, AFz, F1, F5, F6, AF8, AF4, F2), where it showed a higher response than first harmonic responses, t(41) = −3.599, p < .001, and also compared to the frontal 4Hz activity for the no-flicker and omission condition t(40) = −3.606, p < .001.

SNR-normalized frequency spectra and topographies for periodic stimulation frequencies (from top to bottom: 2, 4, 6, 8, 10, 15, 20, and 30 Hz).

The left panel shows the SNR-spectrogram for each of the eight discrete stimulation frequencies. The right panel shows the topographic maps of the first, second and third harmonic, and at 4 Hz.

SNR-normalized frequency spectra for the no-flicker and omission condition.

(A) Frequency responses to trials in which an image was presented without flickering. (B) Responses following the presentation of a blank screen. One-sample t-tests against an SNR = 1 revealed significant increases in SNR-corrected power at 4 Hz and 4.5 Hz for the no flicker condition: t(40) = 3.287, p = .001 and t(40) = 4.874, p < .001, respectively. In contrast, no significant effects were observed for the omission condition: t(40) = 1.460, p = .08 (4 Hz) and t(40) = 1.306, p = .10 (4.5 Hz). Note: N = 41 participants were included in these analyses, as one infant did not contribute sufficient trials.