Figure 1.
‘On’ and ‘off’ sequences were alternatively presented. ‘On’ sequences consisted of checkerboard reversal sweeps of 22 stimuli. The stimulus onset asynchrony (SOA) between pattern reversals was either constant (isochronic conditions) or variable (jitter conditions). The presentation of isochronic and jittered sequences at different rates was randomized. ‘Off’ sequences consisted of grey stimuli lasting 2 to 5 seconds. Note: checkerboards comprised 8 by 8 checks; the figure shows 2 by 2 checks for illustrative purposes.
Table 1.
Experiment 1.
Table 2.
Experiment 2.
Figure 2.
Procedure to synthesize steady-state responses.
Synthetic data for each subject and stimulation rate were obtained by linear addition of the corresponding transient response template. A, The template was time-shifted at its corresponding SOA to create 22 synthetic trials. Here, we show as an example the transient template for the 12.5 rev/s condition of Experiment 1, time-shifted every 80 ms. B, The first five trials and the last one were discarded as for the recorded data. The remaining 16 synthetic trials were linearly added to simulate the response to one stimulation sequence. The triangles indicate stimulus onset. C, The synthetic signal was segmented into stimulus-locked epochs that were averaged. The epoch lasting from 300 ms pre-stimulus to 500 ms post-stimulus was extracted to be compared with its corresponding recorded waveform.
Figure 3.
Recorded isochronic waveforms (Experiment 1).
A, Grand-average waveforms (800 ms time window, including 300 ms pre-stimulus activity) for the five isochronic conditions. The typical deflections of the pattern reversal ERP (N75 at 70–90 ms, P100 at 80–120 ms and N135 at 120–180 ms) are indicated in the grand-average for the 2.7 rev/s condition. The triangles placed below the waveforms indicate the timing of stimuli appearance; the largest triangle represents the zero time point. B, Power spectral densities of the grand-average waveforms for the five isochronic conditions.
Figure 4.
Templates of transient responses (Experiment 1).
Grand-average waveforms (800 ms time window; 300 ms pre-stimulus) of the transient templates used to create the synthetic data at the five different rates. The transient templates correspond to the stimulus-locked average of the jitter conditions.
Figure 5.
Synthetic waveforms obtained from linear superposition of transient templates (Experiment 1).
The synthetic data are shown in black; the recorded data have been overlaid in grey for comparison. A, Grand-average waveforms (800 ms time window; 300 ms pre-stimulus) of the synthetic data at the five different rates. The triangles indicate stimuli onset; the largest triangle represents the zero time point. B, Power spectral densities of the grand-average waveforms.
Figure 6.
Grand-average waveforms at the Oz lead for both isochronic and jitter conditions (Experiment 2).
A, Recorded waveforms (800 ms time window; 300 ms pre-stimulus) for the isochronic stimulation conditions, including one stimulation rate characteristic of transient ERPs (2.5 rev/s) and nine typical SSR reversal rates ranging from 7.7 to 20 rev/s. The triangles indicate stimuli onset; the largest triangle represents the zero time point. B, Grand-average waveforms of the corresponding jitter conditions are shown in solid line. Reminiscent activity from preceding and forthcoming stimuli has been marked by arrows. The Gaussian-modulated waveforms used as transient templates in the synthesis procedure are shown in dashed line.
Figure 7.
Synthetic waveforms obtained from linear superposition of transient templates (Experiment 2).
The synthetic waveforms for the Oz lead at each stimulation rate are shown in black; the recorded data have been overlaid in grey for comparison. A, Grand-average waveforms of the synthetic data at the ten different rates studied. The triangles indicate stimuli onset; the largest triangle represents the zero time point. B, Power spectral densities of the grand-average waveforms.
Figure 8.
Voltage topography for both recorded and synthetic data (Experiment 2).
Scalp voltage topographies for the dominant frequency of each condition (rows) and phase angle (columns) from 0° to 180°. Recorded topographies are presented in the solid boxes; synthetic topographies, in the dashed boxes. The time points corresponding to each phase angle are indicated by its corresponding colour in the time courses of the frequency amplitude at Oz shown on the left side. Phase angles are referred to stimulus onset (i.e. 0° phase angle corresponds to 0 ms).
Figure 9.
Simulation of the influence of both global and local stimulation frequency on the shape of the average ERP.
A, Example of the response elicited by a jittered stimulation train in both the oscillatory entrainment and the superposition scenarios. B, Average waveforms for different stimulation frequencies. The solid bars along the x axis reflect the peak-to-peak latency. C, Average waveforms for trials grouped by their immediately past SOA. D, Variation in the peak-to-peak latency in both scenarios, as a function of both mean stimulation rate and local frequency.
Figure 10.
Effect of global and local stimulation frequency on the shape of recorded ERPs.
A, Grand-average ERPs for the jitter conditions of Experiment 2. The solid bars along the horizontal axed indicate the peak-to-peak (N75-P100) latency. B, Grand-average ERPs for trials grouped by their immediately past SOA at each stimulation rate. C, Peak-to-peak latencies as a function of both global and local stimulation frequency. The error bars indicate the standard error of the mean across subjects.
Figure 11.
Absence of additional activity beyond the last stimulus of the sequence.
The figure shows the grand-average ERPs time-locked to the last stimulus of the train for both isochronic (red lines) and jitter (blue lines) conditions from Experiment 2. The shaded area represents the standard error of the mean across subjects. Triangles indicate the onset of the last stimulus. The time window corresponding to the P100 component has been marked by the first column of coloured boxes (from 85 to 110 ms after stimulus onset; red: isochronic conditions; blue: jitter conditions). The time window corresponding to a potential additional response has been marked by the second column of coloured boxes, with different latency depending on stimulation rate. The asterisks indicate average responses significantly higher than 0 (*p<.05; **p<.01; ***p<.001); whereas the absence of asterisks indicates no significant responses (p>.05).