Summary: Using depth electrodes implanted in patients with epilepsy, researchers tracked the rapid spread of neural activity underlying exogenous attention—the automatic reorientation of focus triggered by external events. By recording from roughly 1,400 intracortical contacts across 28 individuals, the study reveals how attention unfolds across a continuum of cortical networks, moving from early visual processing to response preparation and execution.
The findings identify three sequentially activated neural networks and clarify the neural basis of the attention phenomenon known as “inhibition of return,” a mechanism that helps the brain deprioritize already-examined regions of space. This work offers new insights into attention dynamics and has potential implications for improving rehabilitation strategies for stroke survivors and other patients with attentional deficits.
Key Facts:
- Depth Electrode Precision: Intracortical electrodes implanted for epilepsy monitoring provided high-resolution recordings from about 1,400 brain contacts, giving a rare, detailed view of attention-related neural activity across the cortex.
- Attention Unfolds as a Gradient: Data show a spatiotemporal progression of activity—from parieto-occipital visual areas toward frontal regions associated with decision and action—indicating a continuous flow from perception to response.
- Inhibition of Return Explained: The study identifies neural correlates of inhibition of return, a filter that reduces attention to previously inspected locations and supports efficient visual exploration; understanding this process may guide treatments for attention impairments.
Source: Paris Brain Institute
In today’s environment of nonstop notifications, ads, and streams of information, our attention is constantly challenged. How does the brain decide what captures our focus, and why do some external events so readily pull our attention away?
“Exogenous attention is an automatic process: a salient stimulus can pull our focus without conscious intent. For example, when a colleague passes by, our gaze can involuntarily leave the screen,” explains Tal Seidel Malkinson (University of Lorraine), formerly a postdoctoral researcher at the Paris Brain Institute and now a professor and neuroscientist.
Despite being familiar to anyone who tries to concentrate, the brain mechanisms that drive this involuntary shift in attention have remained incompletely understood, in part because common imaging tools trade off spatial for temporal resolution.
Functional MRI provides good spatial maps but limited temporal detail; EEG offers millisecond timing but coarser localization. To capture both the large-scale and fast dynamics of attention, the team needed direct electrical recordings from neurons across the cortical surface and deep structures.
Direct recordings at neuronal scale
To achieve this, the researchers studied 28 patients undergoing surgical evaluation for drug-resistant epilepsy who had depth electrodes implanted as part of clinical care. Electrode placement was individualized, collectively covering approximately 1,400 intracranial contacts and enabling precise measurement of neural activity while participants performed attention tasks.
In the task, participants fixated on a central cross between two peripheral boxes. Targets appeared unpredictably in one box, and subjects indicated detection by pressing a button. Before some targets, a brief peripheral cue either correctly indicated the upcoming target location (valid cue) or misled attention toward the opposite side (invalid cue).
This classic cueing design allows researchers to measure when attention is drawn to a location, how quickly it shifts, and how the brain groups visual events either as connected or separate occurrences, as explained by co-author Paolo Bartolomeo (Inserm).
To make sense of the complex intracranial data, Tal Seidel Malkinson and Jacobo Sitt (Inserm) developed an unsupervised algorithm that clusters electrodes based on similar temporal activity patterns. This data-driven approach reveals large-scale network dynamics without imposing prior theoretical assumptions.
A cortical gradient from vision to action
Analysis revealed three neural clusters activated in sequence from the back of the brain to the front. Early activity in parieto-occipital regions corresponded to visual processing. Later responses in frontal regions aligned with behavioural preparation and the timing of the button press. Attention therefore appears to emerge as a continuous bridge linking perception and action across cortical gradients.
“We observed a continuum of activity: visual attributes shape early responses at one end of the gradient, while the other end reflects the upcoming motor response,” Malkinson says. “Attention functions as the link between these poles—connecting what we see to how we act.”
The high-resolution temporal and spatial profile of these networks allowed the team to pinpoint the neural signature of inhibition of return: slower reaction times when targets reappear in locations that were recently inspected. This mechanism likely helps the brain prioritize unexplored areas during visual search, improving efficiency.
“Inhibition of return supports effective exploration by reducing attention to familiar locations,” the researchers note. The phenomenon is often disrupted in stroke patients, and a better mechanistic understanding of attention could inform therapeutic approaches to restore or compensate for such deficits.
About this visual neuroscience and attention research news
Author: Marie Simon
Source: Paris Brain Institute
Contact: Marie Simon – Paris Brain Institute
Image: The image is credited to Neuroscience News
Original Research: Open access. “Intracortical recordings reveal Vision-to-Action cortical gradients driving exogenous attention” by Tal Seidel Malkinson et al., published in Nature Communications.
Abstract
Intracortical recordings reveal Vision-to-Action cortical gradients driving exogenous attention
Exogenous attention makes salient external stimuli stand out in a visual scene and is critical for adaptive behaviour. How such attention-capturing events modulate large-scale human brain processing has been unclear. By analyzing activity from 1,403 intracortical contacts implanted in 28 participants performing an exogenous attention task, the authors mapped a spatiotemporal gradient of three neural clusters across the cortex.
The timing, anatomical location, and task relevance of attentional events defined a hierarchy of timescales and mapped onto cortical gradients. Visual features drove activity at one end of the gradient, while response-related signals dominated the opposite end; attentional effects emerged where these signals intersected. These results challenge strictly modular, multi-step models of attention and suggest that frontoparietal networks that treat sequential inputs at the same location as distinct events contribute to exogenous attention phenomena such as inhibition of return.