Phasic modulation of attentional rhythmic sampling according to task demands

doi:10.21203/rs.3.rs-4692989/v1

Download PDF

Article

Phasic modulation of attentional rhythmic sampling according to task demands

https://doi.org/10.21203/rs.3.rs-4692989/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Attention is considered to periodically allocate cognitive resources to task-relevant stimuli, thereby sequentially sampling multiple stimuli in the temporal domain. However, whether the temporal dynamics of attentional sampling are modulated by the task at hand has yet to be fully elucidated. This study explores the dynamics of attentional sampling and the task modulation of these dynamics using high-frequency steady-state visual evoked potential (SSVEP). Participants were asked to select one of two visual stimuli or integrate both as a single target, with stimuli tagged at 42 Hz and 44 Hz. Analysis of SSVEP envelopes revealed the attention periodically samples at theta band (4-6 Hz) in both tasks. Notably, the phase relationship between the SSVEP envelopes differed between tasks, with distinct phases in the selection task and similar phases in the integration task. This phase modulation of the task was mediated by neural oscillations in the right frontoparietal areas. The phase difference between SSVEP envelopes correlated with behavioral performance in the integration task but not in the selection task. These results demonstrate a flexible and task-dependent mechanism for attentional rhythmic sampling, in which neural oscillations align stimuli processing according to task demands.

Biological sciences/Psychology

Biological sciences/Neuroscience/Cognitive neuroscience/Attention

Visual attention

rhythmic sampling

task modulation

phase modulation

Attention allocates limited cognitive resources to task-relevant stimuli to enhance task performance ^1,2. While attention was once thought to be continuously focused on a selected target ³, recent behavioral and neuroscience research has found that attention is more like a “blinking spotlight” that periodically samples multiple stimuli ⁴. Behavioral dense sampling and neurophysiological research have found that attention samples at a rate of about 8 Hz when faced with a single target. In contrast, when faced with two targets, the sampling rate changes to about 4 Hz, suggesting that attentional sampling dwells on each target for a relatively fixed amount of time ^5–9. However, other studies have found that attentional sampling is flexible and adaptive. When the number of targets is increased to four, the sampling rate is accelerated to ensure a fixed oscillation period and thus maintain perceptual continuity ¹⁰. Periodic attentional sampling also accelerates in complex visual search tasks ¹¹. However, tasks in the real world are complex, and in some cases, a single target is selected from multiple candidate stimuli, whereas in other cases, attention must integrate multiple stimuli into a single target for detection. Whether the sampling dynamics of attention are similar in these cases remains unclear.

Neural oscillations in the brain play an important role in the dynamic sampling of attention. The frequency of neural oscillations, especially the individual alpha frequency, represents the rate of attentional sampling ^12,13. When the frequency is high, the period of attentional sampling is short, while when the frequency is low, the period of attentional sampling is long. The phase of neural oscillations, particularly in the theta and alpha bands, is thought to play an important role in target selection, distractor suppression, attentional shifts, and saccades ^14–16. Multiple stimuli can be processed in different phases of an oscillatory cycle, with task-relevant stimuli being assigned to excitatory phases and task-irrelevant stimuli to inhibitory phases ^4,17. The excitatory phases are used for attentional engagement, while the inhibitory phases promote attentional shift and saccade ^4,18. The effects of tasks on attentional sampling are also reflected in neural oscillations. Temporal integration/segmentation tasks affect the alpha frequency of neural oscillations and thus, the temporal resolution of attentional sampling ¹³. In the spatial domain, however, it has not been investigated whether single-stimulus selection and multi-stimulus integration lead to different attentional sampling dynamics and associated neural oscillations.

When investigating the attentional sampling of multiple stimuli, a technical difficulty is to isolate and compare the neural responses to each stimulus from the overall neural responses elicited by simultaneously presented multiple stimuli. The steady-state visual evoked potential (SSVEP) provides an excellent solution to this problem by directly measuring the neural response synchronized with the flicker frequency of the stimulus, thereby isolating the response to different stimuli ^19,20. At the same time, SSVEP power is sensitive to attention, and when attention is directed to a particular stimulus, the corresponding SSVEP power is elevated ^21,22. Previous SSVEP studies have mostly used low frequencies, which interfere with the prevalent low-frequency oscillations in the brain, such as theta, alpha, and beta oscillations. Therefore, it has been difficult to investigate the interaction between the changes in SSVEP and the endogenous low-frequency oscillations in the brain. In the present study, we used high-frequency SSVEP ²³ to tag the presented stimuli to investigate the attentional sampling of multiple stimuli, the task’s effects on the sampling dynamics, and its neural oscillatory mechanism.

In the present study, we asked participants to select one of the stimuli presented on either side of the fixation point as the target or to integrate the two stimuli as a single target, and we tagged the stimuli with high-frequency flicker (42 Hz on the left and 44 Hz on the right). By analyzing the temporal dynamics of the SSVEP envelopes, we investigated the dynamics of attentional sampling over the two stimuli and the effect of the task on them. The results showed that the envelopes of the SSVEPs for both stimuli exhibited fluctuations in the theta band (4.4 Hz and 6.2 Hz), either in the single-stimulus selection task or in the two-stimulus integration task. Critically, the two SSVEP envelopes showed a significant phase difference during the selection task, indicating different temporal sampling for the two stimuli. In contrast, the integration task, which required spatial integration, showed synchronized sampling as reflected by a smaller phase difference between the two envelopes. Furthermore, by analyzing the coupling between SSVEP amplitude and EEG phase, we identified the right frontoparietal areas responsible for modulating these sampling dynamics through phase modulation. Finally, we observed a significant correlation between the observed phase differences and behavioral performance in the integration task. This study identifies a flexible, task-dependent mechanism for attentional sampling, whereby attention aligns different stimuli to be processed at different times when selecting a single stimulus and aligns different stimuli to be processed at nearly same time when integrating multiple stimuli.

High-frequency tagging of visual stimuli in different task demands

In this experiment, participants performed two tasks in two blocks, either selecting a single stimulus or integrating two stimuli as target (Fig. 1A). At the beginning of the experiment, participants were presented with two 2*2 squares in the left and right visual fields. After 500 ms, the luminance of one of the squares began to vary with a sine wave, while the luminance of the other square also began to vary with another sine wave 200 ms later (Fig. 1B). The sine wave in the left visual field was 42 Hz, while the sine wave in the right visual field was 44 Hz. This asynchronous presentation served as a resetting cue to direct attention to the stimulus whose luminance changed first at the onset of the cue ²⁴. After a random interval of 1400–1700 ms, two arrows were presented with each in a random square of the left and right visual fields. In the selection task, participants judged whether the vertical arrow was up or down, whereas, in the integration task, participants judged whether the integrated direction of the two arrows was primary or secondary diagonal (Fig. 1C). Specifically, left-up integration and right-down integration were primary diagonals, whereas left-down integration and right-up integration were secondary diagonals.

We extracted the data from 1700 ms after the cue onset in each trial for further analysis. In both the selection and integration tasks, we obtained event-related potentials (ERPs, Fig. 1D&E) after averaging the left-first-vary and right-first-vary trials, respectively. The results showed that the ERPs contained significant high-frequency oscillations, indicating that our high-frequency stimuli elicited significant high-frequency responses.

We performed FFT analysis on ERPs to further validate the SSVEP for high-frequency stimuli in the present study. The results showed significant 42 Hz and 44 Hz responses in both the selection and integration tasks (Fig. 2A&B). Furthermore, the 42 Hz response was mainly observed in the right occipital area, while the 44 Hz response was mainly observed in the left occipital area, consistent with our stimulus manipulation. Moreover, we computed the power of SSVEP after the cue presentation and compared it statistically with the baseline (i.e., 200 ms before the cue). The SSVEP power was found to be significantly higher than the baseline in both the selection task and the integration task (all p < 0.05, all Cohen’s d > 0.46; Fig. 2C&D). This further suggests that our experimental manipulations obtained high-quality SSVEP signals.

Rhythmic sampling of visual stimuli in theta band

We investigated the dynamic allocation of attention between the two stimuli by analyzing the temporal dynamics of the SSVEP amplitude. First, we extracted the envelope of the SSVEP amplitude in the occipital electrodes contralateral to the stimulus, i.e., 44 Hz SSVEP in the PO3 electrode and 42 Hz SSVEP in the PO4 electrode, and performed an FFT analysis on the envelope. To remove the effect of aperiodic fluctuations on the analysis of periodic oscillations, we used the fooof algorithm ²⁵ to remove the aperiodic components from the spectrum. The results showed that the envelope of PO3 fluctuated periodically at 6.2 Hz, while the envelope of PO4 fluctuated periodically at 4.4 Hz in both the selection task (Fig. 3A) and the integration task (Fig. 3B). Notably, the 4.4 Hz envelope oscillation corresponding to the left stimulus (42 Hz) appeared in the right occipital area, whereas the 6.2 Hz envelope oscillation corresponding to the right stimulus (44 Hz) appeared in the left occipital area, in both the selection task (Fig. 3C) and the integration task (Fig. 3D), consistent with the topographic distribution of the SSVEP power. These results indicate that attentional sampling is periodic for both left and right stimuli, resulting in periodic fluctuations of SSVEP envelopes.

Previous studies of attentional oscillations based on behavioral dense sampling have found that there is a fixed phase difference in the periodic sampling of two spatial locations ^9,26. In the present study, we similarly analyzed the relationship between attentional sampling of left and right stimuli by analyzing the phase relationship between the two SSVEP envelope oscillations. In analyzing the phase relationship, we chose an intermediate frequency of 5.2 Hz between 4.4 and 6.2 Hz. In the selection task (Fig. 4A), when the left side (42 Hz) was first to vary, the phase of the 42 Hz SSVEP envelope was significantly clustered around 126° (z = 19.99, p < 0.001), whereas the phase of the 44 Hz SSVEP envelope was significantly clustered around 76° (z = 8.86, p < 0.001). There was a significant phase difference between the SSVEP envelopes of the left and right stimuli, clustered around 48° (z = 9.12, p < 0.001). In contrast, in the right (44 Hz) first-to-vary condition, the phase of the 44 Hz SSVEP envelope was significantly clustered around 120° (z = 11.19, p < 0.001), whereas the phase of the 42 Hz SSVEP envelope was significantly clustered around 84° (z = 19.20, p < 0.001). There was also a significant phase difference between the SSVEP envelopes of the right and left stimuli, clustered around 36° (z = 9.70, p < 0.001). These results suggest that the resetting cue resulted in a temporal difference between the attentional periodic sampling of the two stimuli. They validate the results of previous behavioral oscillation studies ²⁴ and confirm that the cue manipulation was effective in the present experiment.

In the integration task (Fig. 4B), when the left side (42 Hz) was first to vary, the phase of the 42 Hz SSVEP envelope was significantly clustered around 97° (z = 19.98, p < 0.001), while the phase of the 44 Hz SSVEP envelope was significantly clustered around 109° (z = 7.86, p < 0.001). There was a significant phase difference between the left and right SSVEP envelopes, clustered around − 12° (z = 8.16, p < 0.001). In contrast, when the right side (44 Hz) was first to vary, the phase of the 44 Hz SSVEP envelope was significantly clustered around 110° (z = 9.70, p < 0.001), while the phase of the 42 Hz SSVEP envelope was significantly clustered around 122° (z = 19.88, p < 0.001). There was a significant phase difference between the right and left SSVEP envelopes, clustered around − 12° (z = 9.79, p < 0.001). Notably, the phase differences for both the left-first and right-first conditions were not significantly different from 0° (v = 12.82, p < 0.001 for the left-first condition; v = 14.03, p < 0.001 for the right-first condition). By combining the left-first and right-first conditions, we compared the SSVEP envelope phase differences of the first-vary and the last-vary stimuli between the selection task and the integration task. It was found that there was a significant difference in the phase differences between the selection task and the integration task (F(2, 19) = 7.17, p = 0.005). This suggests that the task influences the sampling dynamics of attention through temporal modulation. In the integration task, attention sampled the two stimuli almost synchronously, whereas in the selection task, attention sampled the two stimuli at different times.

Rhythmic sampling of visual attention is mediated by the phase of neural oscillation

Since the task modulates the phase of the periodic sampling of attention, this suggests an endogenous mechanism that controls the sampling dynamics of attention. In this section, we explore this endogenous mechanism by examining the coupling between SSVEP amplitude and the phase of neural oscillations in the brain. The coupling between the SSVEP amplitude and the phase of the EEG is equivalent to the coherence between the phase of the SSVEP amplitude oscillations and the phase of the EEG. We calculated the coherence between the phase of the EEG at all electrodes and the phase of the SSVEP amplitude oscillations at PO3 and PO4. As shown in Fig. 5A, in the selection task, the coupling appeared mainly on electrodes in the right parietal and frontal areas, and was significant in the right parietal area in the left-first condition. As shown in Fig. 5B, in the integration task, coupling appeared mainly on electrodes in the middle parietal and frontal areas, and was significant in both the left-first condition and the right-first condition. This result suggests that a control mechanism in frontoparietal areas influences the periodic sampling of attention by phase modulation of neural oscillations.

Behavioral correlates of phase modulation

We further investigated the behavioral correlates of the oscillatory phase difference between the SSVEP envelopes by examining the correlation between phase difference and participants’ reaction time in the selection and integration tasks, respectively. We also examined the correlation between absolute phase difference and reaction time. The results showed that in the selection task, both phase difference (r = -0.04, p = 0.875) and absolute phase difference (r = -0.158, p = 0.522) were not significantly correlated with reaction time. In the integration task, phase difference was not significantly correlated with reaction time (r = 0.13, p = 0.577), but there was a significant positive correlation between absolute phase difference and reaction time (r = 0.52, p = 0.015). This result may indicate that in the selection task, the magnitude of the phase difference had little effect on participants’ selection of one of the two targets, but when integrating the two targets, the smaller the phase difference, the better the integration.

This study investigated the dynamics of attentional sampling when participants selected one of two stimuli as a target or integrated them into a single target. By employing high-frequency SSVEP and analyzing the temporal dynamics of the SSVEP envelopes, we found that attention periodically sampled both stimuli in the theta band (around 4–6 Hz). Critically, the task modulated the phase relationship between these periodic dynamics. During the selection task, attention sampled the two stimuli at distinct phases, suggesting a focus on each stimulus at different times. Conversely, during the integration task, attention sampled the stimuli with minimal phase difference, suggesting a more synchronized processing of both stimuli. The sampling profiles were mediated by neural oscillations in the right frontoparietal brain regions through phase-amplitude coupling. The phase difference between the sampling rhythms of the two stimuli influenced the behavioral performance in the integration task. These findings support a task-dependent temporal modulation of attentional periodic sampling that is mediated by the phase of neural oscillations in the brain.

This study used high-frequency tagging to investigate the dynamics of attentional sampling of stimuli and showed a coupling between the phase of low-frequency EEG signals and the amplitude of high-frequency SSVEP. Previous studies have shown that the early visual cortex can generate SSVEPs to flickering stimuli up to 100 Hz ²⁷ and that gamma-band tagging is independent of endogenous alpha-band neural oscillations ²³. In addition, the neural response of the SSVEP was weaker at higher frequencies, even for stimuli of the same intensity ²³. Combining these factors, stimulus flicker in the low gamma band (42 Hz and 44 Hz) was chosen for tagging in the present study. This allowed both a relatively high signal-to-noise ratio for SSVEP and the study of the interaction between changes in the SSVEP envelope and endogenous low-frequency neural oscillations. Spatial attention affects the amplitude of the SSVEP, and in particular, attentional engagement increases the amplitude of the SSVEP in the gamma band ^22,23, allowing fluctuations in SSVEP amplitude in the present study to reflect dynamic changes in attentional allocation. The results of the present study provide a feasible solution for tagging perceptual and attentional processing and demonstrate that high-frequency tagging can be used to study the interaction between stimulus responses and endogenous low-frequency neural oscillations.

The coupling analysis between SSVEP amplitude and EEG phase revealed right frontoparietal areas as potential modulators of attentional sampling dynamics according to task demands. Theta phase in frontoparietal areas has been previously suggested to coordinate two rhythmically alternating attentional states that are linked to attentional engagement, attentional shifts, and saccade generation ^15,16. Our findings go further step in demonstrating that these brain areas might mediate the phase of theta-band oscillations according to the task demand, leading to a temporally distinct sampling of stimuli during the selection task versus synchronized sampling during the integration task. The results of the current study are also consistent with LJ-multiplexing proposed by Lisman and Jensen²⁸. According to this theory, the neuronal population representing a given stimulus fires together within a single gamma cycle, with the inhibitory portion of the gamma cycle yielding a temporal separation before the activation of the next neuronal population representing another stimulus. The theta/alpha oscillation structures the order of activation of these neuronal populations because as the oscillatory inhibition decays on each cycle, more strongly activated populations are activated earlier ^17,29. In our results, when the task requires selecting one of two stimuli as a target, the responses to the two stimuli are separated in the temporal domain, suggesting that the two stimuli are sampled sequentially by attention, whereas when the task requires integrating the two stimuli as a single target, the responses to the two stimuli tend to be synchronized in time, suggesting that the two stimuli are sampled synchronously by attention.

Interestingly, although the cue manipulation resulted in a stable phase difference between the SSVEP envelopes of the two stimuli in the selection task, the magnitude of this phase difference was not correlated with behavioral performance. This may be because the phase difference between the two SSVEP envelopes indicates that attention segmented the two stimuli in the temporal domain but did not enhance the processing of the individual stimuli. In the integration task, the phase difference between the two SSVEP envelopes was close to 0, indicating that attention modulated the temporal dynamics of the SSVEP according to the task. Meanwhile, the absolute magnitude of the phase difference correlated with behavioral performance. This suggests that a smaller phase difference, reflecting more synchronized processing, might be beneficial for integrating the two stimuli. This finding aligns with the notion that successful integration likely requires processing the two stimuli at similar points in time ¹³.

While this study sheds light on the task-dependent modulation of attentional sampling, some limitations need to be addressed in future research. First, the present study focused on the behavioral performance and neural oscillatory correlates of attentional sampling in a simple two-stimulus paradigm. Future studies could explore how attentional sampling dynamics are modulated in more complex real-world scenes with multiple stimuli. Second, the present study used high-frequency SSVEP to isolate the neural response to each stimulus. However, the spatial resolution of the EEG signal in this study is blurred by the volume conduction effect, and future research combining the SSVEP with techniques like MEG and sEEG could pinpoint the specific brain regions responsible for these phase modulations.

In conclusion, this study demonstrates that attentional sampling is flexible and adapts to task demands with distinct temporal relationships between the sampling of two stimuli during selection versus integration tasks. The findings suggest that right frontoparietal areas might mediate the phase of theta-band neural oscillations to control the temporal dynamics of attentional sampling, ultimately optimizing processing for specific task goals.

Participants

Twenty-one adults were recruited from Hangzhou Normal University to participate in the experiment. The average age of the subjects was 22 years old (9 males). All subjects were right-handed, had normal vision or corrected-to-normal vision, and had no history of psychological or psychiatric illness. All subjects provided written informed consent prior to the start of the experiment, which was approved by the Research Ethics Committee at Hangzhou Normal University

Apparatus and stimuli

We used MATLAB (version R2021a, The MathWorks) and Psychotoolbox-3 extensions ^30,31 to generate and display visual stimuli and record behavioral responses. The visual stimuli were presented on a QD-OLED monitor (Alienware, Dell) with a 1920 × 1080 spatial resolution and a 144 Hz refresh rate. Electroencephalography (EEG) was recorded in a dark room shielded from sound and electromagnetic signals. The participants were comfortably seated at 57 cm from the screen with their heads stabilized on a chin rest.

In this experiment, we presented participants with a 2*2 square (length of side: 24.65°) in each of the left and right visual fields on a black background (at minimum luminance 0 cd/m²). The line width of the square was 10 pixels. We used a frequency-tagging technique to record the brain’s response to the stimuli in left and right visual fields. The luminance of the squares presented in the left and right visual fields varied sinusoidally between the minimum and maximum luminance (266.73 cd/m²) of the screen at 42 Hz and 44 Hz, respectively. The targets were two arrows (266.73 cd/m²) with each presented in a random square of the left and right visual fields. One arrow randomly pointed up or down, and the other randomly pointed left or right.

Design and procedure

Trials began with the presentation of the squares in the left and right visual fields for 500 ms, after which the square on a random side began to flash, and then 200 ms later, the square on the other side also began to flash. This asynchronous flashing served as a resetting cue for attentional processing, allowing attention to be directed to the side that flashed first after the presentation of the resetting cue ²⁴. After a random interval of 1400 to 1700 ms, both arrows were presented simultaneously for 100 ms in the squares of the left and right visual field as targets.

In two blocks, participants were required to complete two separate tasks. In the selection task, participants judged whether the vertical arrows were pointing up or down. In the integration task, participants judged whether the combined direction of the two arrows was the principal diagonal or the secondary diagonal. Explicitly, the combined direction of the left and up arrows and the combined direction of the right and down arrows was the primary diagonal, and the combined direction of the left and down arrows and the combined direction of the right and up arrows was the secondary diagonal. To avoid the effect of symmetric visual fields on target integration, the two targets were presented in random squares in the left and right visual fields. Throughout the experiment, participants were asked to fixate on the dot in the center of the screen and to respond within 2000 ms of the target presentation.

The order of presentation of the two tasks was counterbalanced between participants, and each task lasted approximately 20 minutes. Before the start of each task, participants were required to complete 16 trials of practice to get familiar with the experimental task.

EEG recording and pre-processing

EEG data were recorded at a sampling rate of 1000 Hz using a 64-channel EasyCap and two BrainAmp amplifiers (Brain Products, Germany). The vertical electrooculogram (EOG) was recorded by an additional electrode placed below the right eye. FCz was used as the reference electrode during recording. The impedance of all electrodes was kept below 5 kΩ. Offline preprocessing was carried out using Fieldtrip toolbox ³². The EEG data were re-referenced to the average value of TP9 and TP10 and band-pass filtered between 0.1 and 50 Hz using a Butterworth IIR filter with the order of 2. Independent component analysis (ICA) was performed to remove eye-movement and other artifactual components. No further trial rejection was executed. Data were resampled to 500 Hz and epoched from − 200 ms to 1700 ms relative to the onset of the resetting cue.

Spectrum analysis of SSVEP

Evoked activities of baseline (-200-0 ms) and SSVEP (0-1700 ms) were computed respectively by averaging EEG epochs for each condition and each participant. FFT was applied to the evoked activities after applying a 10 s zero-padding and a Hanning taper to calculate the power spectrum for each channel from 36 to 48 Hz with a 1 Hz step.

Spectrum analysis of the envelope of SSVEP

To calculate the power spectrum of the envelope of the SSVEP, we first filtered the EEG of each trial through a bandpass filter to obtain SSVEPs at 42 Hz (41.5–42.5 Hz) and 44 Hz (43.5–44.5 Hz). Next, we calculated the envelope of the SSVEPs using the envelop function of MATLAB (version R2023b, The MathWorks), and averaged them across trials. FFT was then applied to the trial-averaged envelope after applying a 10 s zero-padding and a Hanning taper to calculate the power spectrum for each channel from 2 to 10 Hz with a 0.2 Hz step. Finally, we used Fieldtrip’s fooof function ²⁵ to extract the periodic oscillations in the spectrum and remove the aperiodic components. To examine the phase relationship between the two SSVEP envelopes in the left and right visual field, we extracted the phase of 5.2 Hz, the mid-frequency between 4.4 Hz and 6.2 Hz.

EEG phase and SSVEP coupling

We aimed to calculate the coupling between the SSVEP amplitude and the phase of the EEG, which is equivalent to the coherence between the phase of the SSVEP envelope and the phase of the EEG. We used the same spectral analysis for EEG as for the SSVEP envelope and calculated the coherence between the phase of the SSVEP envelope and the phase of the EEG signal for the left-first and right-first conditions, respectively. Based on the results of the SSVEP envelope oscillations, the phases of the SSVEP envelope were taken from 4.4 Hz at the PO4 electrode and 6.2 Hz at the PO3 electrode, and then we calculated the coherence of the phases of the SSVEP envelope at the two electrodes with the phases of the EEGs at all electrodes.

Behavioral correlates of phase difference

We calculated the phase difference between the SSVEP envelopes for the first-start and last-start flashes in the selection and integration conditions, respectively. After averaging the left-first and right-first trials, we calculated the Pearson correlation between the phase difference and the response time of the correct response. We also examined the Pearson correlation between the absolute value of the phase difference and the response time.

Statistical tests

The statistical comparisons of the powers between baseline and SSVEP were evaluated by paired-sample t-test. Statistical assessment of phase distribution was conducted using CircStat Toolbox ³³. The non-uniformity of the phase and the phase difference were tested using Rayleigh’s test ³⁴. V test was used to evaluate the mean direction of the phase difference against zero ³⁴. The difference of phase differences between the selection and integration tasks was evaluated by Hotelling paired-sample test ³⁴.

We used a permutation test to statistically quantify the significance of the oscillations of the SSVEP envelope. For each trial, we randomly shuffled the envelope signal in time and then used the same FFT and fooof algorithms to compute the periodic oscillations of the surrogate signal. This process was repeated 1000 times to obtain a surrogate distribution of power at each frequency from which a 95% significance threshold (P = 0.05) was obtained.

We also used the permutation test to statistically quantify the phase coherence of the EEG and SSVEP envelopes. In each trial, we randomly shuffled the EEG signals in time and used the same spectral and coherence analyses to compute the phase coherence of the SSVEP envelope and the surrogate EEG signal. This process was repeated 1000 times to obtain a surrogate distribution of coherence values at each electrode. We then calculated the percentile of the real coherence value in the permuted distribution as the significance of the phase coherence.

Pearson correlation was used to evaluate the correlations between phase difference and reaction time.

Acknowledgment

This work was supported by the National Natural Science Foundation of China (32371086), the Zhejiang Provincial Natural Science Foundation of China (LY23C090001), and the Major Project of Philosophy and Social Science Research of the Ministry of Education of China (22JZD044).

Data availability statement

Data and code will be uploaded to the web and made public at the time of publication.

Author contributions

Conceptualization: JJ

Methodology: JJ, QK

Investigation: QK, CQ

Visualization: JJ, QK

Supervision: JJ

Writing—original draft: QK, JJ

Writing—review & editing: JJ, QK

Competing interests

The authors declare that they have no competing interests.

Buschman, T. J. & Kastner, S. From Behavior to Neural Dynamics: An Integrated Theory of Attention. Neuron 88, 127–144 (2015).
Carrasco, M. Visual attention: the past 25 years. Vision Research 51, 1484–525 (2011).
Posner, M. I. Orienting of attention. Quarterly Journal of Experimental Psychology 32, 3–25 (1980).
Fiebelkorn, I. C. & Kastner, S. A Rhythmic Theory of Attention. Trends in Cognitive Sciences 23, 87–101 (2019).
Fiebelkorn, I. C., Saalmann, Y. B. & Kastner, S. Rhythmic Sampling within and between Objects despite Sustained Attention at a Cued Location. Current Biology 23, 2553–2558 (2013).
Jia, J., Fang, F. & Luo, H. Selective spatial attention involves two alpha-band components associated with distinct spatiotemporal and functional characteristics. NeuroImage 199, 228–236 (2019).
Jia, J., Fan, Y. & Luo, H. Alpha-Band Phase Modulates Bottom-up Feature Processing. Cerebral Cortex 32, 1260–1268 (2022).
Re, D., Inbar, M., Richter, C. G. & Landau, A. N. Feature-Based Attention Samples Stimuli Rhythmically. Current Biology 29, 693-699.e4 (2019).
Song, K., Meng, M., Chen, L., Zhou, K. & Luo, H. Behavioral Oscillations in Attention: Rhythmic α Pulses Mediated through θ Band. Journal of Neuroscience 34, 4837–4844 (2014).
Jiang, Y., He, S. & Zhang, J. The adaptive flexibility of rhythmic attentional sampling in attending to multiple targets. Journal of Experimental Psychology: General (2023) doi:10.1037/xge0001468.
Merholz, G., Grabot, L., VanRullen, R. & Dugué, L. Periodic attention operates faster during more complex visual search. Sci Rep 12, 6688 (2022).
Samaha, J. & Postle, B. R. The Speed of Alpha-Band Oscillations Predicts the Temporal Resolution of Visual Perception. Current Biology 25, 2985–2990 (2015).
Wutz, A., Melcher, D. & Samaha, J. Frequency modulation of neural oscillations according to visual task demands. Proceedings of the National Academy of Sciences 115, 1346–1351 (2018).
Busch, N. A., Dubois, J. & VanRullen, R. The Phase of Ongoing EEG Oscillations Predicts Visual Perception. Journal of Neuroscience 29, 7869–7876 (2009).
Fiebelkorn, I. C., Pinsk, M. A. & Kastner, S. A Dynamic Interplay within the Frontoparietal Network Underlies Rhythmic Spatial Attention. Neuron 99, 842-853.e8 (2018).
Helfrich, R. F. et al. Neural Mechanisms of Sustained Attention Are Rhythmic. Neuron 99, 854-865.e5 (2018).
Jensen, O., Gips, B., Bergmann, T. O. & Bonnefond, M. Temporal coding organized by coupled alpha and gamma oscillations prioritize visual processing. Trends in Neurosciences 37, 357–369 (2014).
Lv, X., Cheng, S., Wang, Z. & Jia, J. The dynamics of microsaccade amplitude reflect shifting of covert attention. Consciousness and Cognition 101, 103322 (2022).
Jia, J., Wang, T., Chen, S., Ding, N. & Fang, F. Ensemble size perception: Its neural signature and the role of global interaction over individual items. Neuropsychologia 173, 108290 (2022).
Norcia, A. M., Appelbaum, L. G., Ales, J. M., Cottereau, B. R. & Rossion, B. The steady-state visual evoked potential in vision research: A review. Journal of Vision 15, 4 (2015).
Gordon, N., Tsuchiya, N., Koenig-Robert, R. & Hohwy, J. Expectation and attention increase the integration of top-down and bottom-up signals in perception through different pathways. PLoS Biology 17, e3000233 (2019).
Gulbinaite, R., Roozendaal, D. H. M. & VanRullen, R. Attention differentially modulates the amplitude of resonance frequencies in the visual cortex. NeuroImage 116146 (2019) doi:10.1016/j.neuroimage.2019.116146.
Zhigalov, A., Herring, J. D., Herpers, J., Bergmann, T. O. & Jensen, O. Probing cortical excitability using rapid frequency tagging. NeuroImage 195, 59–66 (2019).
Landau, A. N., Schreyer, H. M., van Pelt, S. & Fries, P. Distributed Attention Is Implemented through Theta-Rhythmic Gamma Modulation. Current Biology 25, 2332–2337 (2015).
Donoghue, T. et al. Parameterizing neural power spectra into periodic and aperiodic components. Nat Neurosci 23, 1655–1665 (2020).
Landau, A. N. & Fries, P. Attention Samples Stimuli Rhythmically. Current Biology 22, 1000–1004 (2012).
Herrmann, C. S. Human EEG responses to 1-100 Hz flicker: resonance phenomena in visual cortex and their potential correlation to cognitive phenomena. Experimental Brain Research 137, 346–353 (2001).
McLelland, D. & VanRullen, R. Theta-Gamma Coding Meets Communication-through-Coherence: Neuronal Oscillatory Multiplexing Theories Reconciled. PLOS Computational Biology 12, e1005162 (2016).
Jensen, O. & Lisman, J. E. Hippocampal sequence-encoding driven by a cortical multi-item working memory buffer. Trends in Neurosciences 28, 67–72 (2005).
Brainard, D. H. The Psychophysics Toolbox. Spatial Vision 10, 433–436 (1997).
Pelli, D. G. The VideoToolbox software for visual psychophysics: transforming numbers into movies. Spatial vision 10, 437–42 (1997).
Oostenveld, R., Fries, P., Maris, E. & Schoffelen, J.-M. FieldTrip: Open Source Software for Advanced Analysis of MEG, EEG, and Invasive Electrophysiological Data. Computational Intelligence and Neuroscience 2011, 1–9 (2011).
Berens, P. CircStat: A MATLAB Toolbox for Circular Statistics. Journal of Statistical Software 1, 1–21 (2009).
Zar, J. H. Biostatistical Analysis (5th Edition). (Pearson Prentice Hall, Upper Saddle River, New Jersey, 2010).

There is NO Competing Interest.

nrreportingsummary1.pdf
Reporting Summary

Download PDF

Version 1

posted

You are reading this latest preprint version

Phasic modulation of attentional rhythmic sampling according to task demands

Status:

Version 1

Abstract

Figures

Introduction

Results

High-frequency tagging of visual stimuli in different task demands

Rhythmic sampling of visual stimuli in theta band

Rhythmic sampling of visual attention is mediated by the phase of neural oscillation

Behavioral correlates of phase modulation

Discussion

Methods

Participants

Apparatus and stimuli

Design and procedure

EEG recording and pre-processing

Spectrum analysis of SSVEP

Spectrum analysis of the envelope of SSVEP

EEG phase and SSVEP coupling

Behavioral correlates of phase difference

Statistical tests

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1