Attention: Temporal, spatial

Talk Session: Tuesday, May 20, 2025, 10:45 am – 12:30 pm, Talk Room 1

Schedule of Events | Search Abstracts | Symposia | Talk Sessions | Poster Sessions

Talk 1, 10:45 am

Voluntary temporal attention enhances informational connectivity across cortical networks

Jiating Zhu¹, Karen Tian^1,2, Marisa Carrasco², Rachel Denison^1,2; ¹Boston University, ²New York University

Motivation: Selective attention enhances communication between brain areas. Whereas spatial attention alters communication through rhythmic synchronization, the impact of temporal attention on communication remains unexplored. Here we investigated whether and how voluntary temporal attention—the goal-directed selection of visual information at specific points in time—dynamically routes stimulus information across cortical networks. Methods: We recorded MEG data in human observers performing orientation discrimination in a two-target temporal cueing task. Observers were instructed with a precue to attend to one of two sequential grating stimuli (T1 and T2) separated by 300 ms, and with a response cue to judge the tilt of the target. To track the flow of stimulus information across the cortex, we constructed a dynamic informational connectivity network using source-reconstructed data. We measured correlations in stimulus orientation decoding accuracy (edges) across all pairs of atlas-based regions (nodes) within sliding time windows. We estimated a region’s closeness centrality as its average inverse correlation distance to all other regions. Higher closeness centrality indicates greater informational connectivity with the rest of the network. Results: We found both early (~100 ms) and late (400-725 ms) modulations of closeness centrality in the informational connectivity network due to temporal attention. These enhancements emerged for both T1 and T2, primarily in the occipital and temporal lobes. We also observed an early occipital motif that recurred periodically in the entorhinal-parahippocampal cortex, but only when a target was temporally attended. The timing of this motif overlapped with the periods of closeness centrality enhancements. Conclusion: The results reveal how temporal attention dynamically routes stimulus information through cortical networks, indicating both early and late selection mechanisms at a network level.

This research was supported by National Institutes of Health National Eye Institute R01 EY019693 to M.C., T32 EY007136 to NYU, and F32 EY025533 to R.D., National Defense Science and Engineering Graduate Fellowship to K.J.T., and startup funding from Boston University to R.D.

Talk 2, 11:00 am

Tracking the neural signatures of internal visual and motor prioritization across space and time

Irene Echeverria-Altuna^1,2, Sage Boettcher², Freek van Ede³, Kate Watkins², Kia Nobre^1,2; ¹Yale University, ²University of Oxford, ³Vrije Universiteit Amsterdam

Visual and motor contents within working memory can be flexibly and dynamically prioritized to guide adaptive behavior. Such prioritization is accompanied by modulations in frequency-specific electroencephalography (EEG) activity. Namely, lateralized posterior alpha-band (8-13 Hz) modulation tracks changes in internal attention to item locations, whereas lateralized central mu/beta-band (8-30 Hz) modulation tracks changes in response-plan prioritization. The limited spatial resolution of EEG, however, precludes resolving the cortico-subcortical networks underlying alpha and mu/beta modulations during internal attention. To investigate and compare the networks controlling internal visual and motor selection associated with these EEG markers, we completed a combined simultaneous fMRI-EEG study. Participants held two items (tilted bars) in working memory. In turn, items appeared on the left or right side and required a left- or right-hand response, with location and hand manipulated orthogonally. In half the trials, an informative retro-cue prompted participants to prioritize one encoded visual item and its associated action plan and, with the passage of time, to shift their focus to the other spatial location and response plan. In the other half of the trials, retro-cues were uninformative, so neither visual nor motor prioritization was possible. The EEG analyses replicated patterns of contralateral alpha modulation for spatial item selection and contralateral mu/beta modulation for motor selection. Analysis of fMRI data revealed the engagement of frontal and parietal areas associated with internal attention control during internal spatial attention and of the corresponding sensorimotor hand representation during internal action selection. Ongoing analyses will probe for cortico-subcortical systems tracking shifts in internal spatial and motor attention. The results promise interesting clues about the sources underlying the modulations of frequency-specific activity that accompany flexible sensory and motor prioritization in working memory.

Talk 3, 11:15 am

Attentional modulation of stimulus-synchronized BOLD oscillations in the human visual cortex

Reebal Rafeh¹, Geoffrey Ngo², Lyle E. Muller³, Ali R. Khan⁴, Ravi S. Menon^4,5, Taylor W. Schmitz², Marieke Mur^6,7; ¹Neuroscience Graduate Program, Western University, ²Department of Physiology and Pharmacology, Schulich School of Medicine & Dentistry, Western University, ³Department of Mathematics, Faculty of Science, Western University, ⁴Department of Medical Biophysics, Schulich School of Medicine & Dentistry, Western University, ⁵Centre for Functional and Metabolic Mapping, Robarts Research Institute, Western University, ⁶Department of Physiology and Pharmacology, Schulich School of Medicine & Dentistry, Western University, ⁷Department of Psychology, Faculty of Social Science, Western University, ⁸Department of Computer Science, Faculty of Science, Western University

Visual cortical neurons synchronize their firing rates to periodic visual stimuli. EEG is commonly used to study directed attention by frequency-tagging brain responses to multiple stimuli oscillating at different frequencies but is limited by its coarse spatial resolution. Here we leverage frequency-tagged fMRI (ft-fMRI) to study the influence of directed attention on the fine-grained spatiotemporal dynamics of competing stimulus-driven visual cortical oscillations. In this 7T fMRI experiment, participants (n=7) distributed their attention to simultaneously presented visual checkerboard stimuli oscillating at 0.125 Hz and 0.2 Hz. Our analysis revealed that distinct populations of visual cortical neurons exhibited either in-phase or anti-phase synchronization with the oscillating stimuli. The spatial topographies of these populations were highly replicable across scan sessions within participants, indicative of a fine-grained map of competitive feature-tuned responses (Dice's coefficient > 0.59; K-S test: p < 10-39). In accordance with this observation, we found that directed attention homogeneously increased the amplitude of anti-phase BOLD oscillations across the visual hierarchy, consistent with a distributed attention-driven suppressive field (Reynolds & Heeger, 2009). In contrast, attentional modulation of in-phase BOLD oscillations increased hierarchically from V1 to hV4, consistent with the effects of target enhancement reported in prior monkey electrophysiology (Moran & Desimone, 1985) and event-related fMRI work (Kastner et al., 1998). Finally, the strength of anti-phase (Wilcoxon test: p = 0.016), but not in-phase (p = 0.16), modulation predicted psychophysical correlates of attentional performance, further highlighting the mechanistic dissociation of attention-driven target enhancement and surround suppression. Together, our findings support the biased competition model of attention, which posits that attention modulates competing neural populations through concurrent mechanisms of enhancement and suppression. ft-fMRI extends the boundaries for research on the neural basis of biased competition in humans by providing a non-invasive method for distinguishing between concurrent stimulus-synchronized in-phase (enhancing) and anti-phase (suppressing) BOLD oscillations.

Talk 4, 11:30 am

Exploring the role of eye movements in the attentional blink phenomenon: insights from naturalistic visual exploration

Shahar Messika¹, Shlomit Yuval-Greenberg^1,2; ¹School of Psychological Sciences, Tel Aviv University, ²Sagol School of Neuroscience, Tel Aviv University

Humans can fixate for extended durations in experimental settings, but this behavior is highly unnatural. The visual system is inherently explorative, and suppressing this drive through fixation tasks may have profound perceptual consequences. We hypothesize that the visual system, due to its explorative nature, is not optimized to perceive sequential foveal targets presented in close temporal proximity. To investigate this, we revisit the well-known attentional blink (AB) phenomenon—a reduced ability to perceive a second target (T2) appearing 200-500ms after an initial target (T1). While various interpretations of the AB exist, its origins remain debated. We propose that the AB arises from the visual system’s intrinsic drive to explore. Specifically, detecting T1 activates an explorative mechanism, shifting attention from the fovea to the periphery in search of the next saccade target, leading to frequent misses of the consecutive foveal target (T2). In two experiments (Total N=47) we tested this hypothesis using a novel AB-like design. Stimuli (digits/letters) appeared sequentially at 10Hz in changing, but predictable, spatial locations along a horizontal path with variability in the vertical-axis. T1 and T2 (letters/digits) were embedded within this sequence, separated by 1-5 lags and participants were asked to report both targets. Participants were not instructed to fixate and naturally tracked the moving stimuli with their gaze. Intermittent trials included a classic AB task with centrally presented stimuli. Results revealed a pronounced AB effect in the foveal condition but a significantly diminished effect in the novel task. These findings suggest that allowing free eye movements and aligning attentional shifts with peripheral exploration reduce the AB effect. We conclude that AB may stem from the unnatural constraints of fixation tasks. This research underscores the importance of dynamic visual exploration in cognition and suggests that classical cognitive phenomena studied under fixation should be re-evaluated in free-viewing conditions.

The study was funded by ISF grant 1960/19 to S-Y.G

Talk 5, 11:45 am

Can you attend broadly in space while attending narrowly in time?: On the generality of attentional breadth

Merve Erdogan¹, Anna C. Nobre¹, Brian Scholl¹; ¹Yale University

A salient aspect of spatial attention is its variable *breadth*: sometimes we select narrowly (e.g. when hunting for lost keys in a cluttered drawer), while other times we select broadly (e.g. when viewing the overall configuration of players on a soccer field). And an analogous dynamic applies across time: sometimes we attend to relatively high-frequency changes (e.g. when listening to fast jazz) while other times we focus on events changing at a slower pace (e.g. waves crashing ashore). How are these different forms of attentional breadth related? For example, can you attend broadly in space while simultaneously attending narrowly in time? One might have no effect on the other. Or attending broadly in one domain might facilitate attending broadly in the other. Or broad vs. narrow attention might draw in part on different resources, such that attending broadly in one domain is easier when attending narrowly in the other. Participants were presented with four types of stimuli at once: (a) visual probes flashed in a relatively narrow ring around fixation (spatially narrow), (b) visual probes flashed in a wider ring relatively far from fixation (spatially broad), (c) repeated-tone probes in a high-frequency auditory stream (temporally narrow), or (d) repeated-tone probes in a lower-frequency stream (temporally broad). Across trials, participants were instructed to attend broadly vs. narrowly in space, and (independently) broadly vs. narrowly in time. As expected, attending to one visual ring impaired performance in the other -- and ditto for the two auditory streams. Critically, there were also cross-dimension interactions: for example, participants were better at focusing spatial attention *broadly* (detecting probes in the wider ring) when their temporal attention was focused *narrowly* (detecting probes in the higher-frequency stream). The interplay between spatial and temporal attention may thus depend on its relative breadth in each domain.

Talk 6, 12:00 pm

Unraveling Learned Distractor Suppression: Insights from Psychophysics and Computational Modeling

Jan Theeuwes¹, Dock Duncan¹, Dirk van Moorselaar¹; ¹Vrije Universiteit Amsterdam

The ability to ignore salient yet irrelevant stimuli is essential to accomplish everyday life activities. Previous research has shown that individuals improve their ability to suppress distracting items through experience; however, the mechanisms underlying this learned suppression remain unclear. The current study employed a psychophysical approach combined with computational modelling investigating how learned spatial suppression affects perception. The results show that items presented at suppressed locations are perceived as less bright than those in non-suppressed areas, suggesting that learned suppression directly affects the perceived saliency of items. To determine how saliency changes impact visual search, computational modelling was used to compare different models of attentional selection. The analysis favored a model in which learned suppression reduces the saliency of objects in suppressed locations during the initial salience computation. As a result, these items are less likely to compete for attentional processing and are therefore less likely to capture attention. To build on these findings, an additional set of experiments focused on high-probability distractor features instead of spatial distributions. The data were analyzed using the same modeling framework to assess whether feature learning operates in a similarly proactive manner. The results indicate that feature learning differs from spatial learning: Instead of a reduction in attentional capture, feature learning results in faster disengagement of attention from high probability distractor features such as frequently encountered colors and shapes. Collectively, these experiments suggest that spatial suppression operates proactively reducing attentional capture at suppressed locations, while feature suppression operates reactively facilitates faster disengagement from frequently encountered features. Together, these findings highlight the distinct mechanisms that drive learned distractor suppression.

ERC advanced grant [833029 - LEARNATTEND] and NWO SSH Open Competition Behaviour and Education grant [406.21.GO.034].