Intracranial Electroencephalography reveals the neural correlates of eye-movement guidance to objects

Arish Alreja, Matthew J. Boring, Michael Ward, Robert Mark Richardson, Taylor J. Abel, Avniel Singh Ghuman · Journal of Vision · 2025

When we move our eyes, it is to look at something: your daughter’s face, the person who just walked into the room, a cup of water, your keys, etc. Yet our understanding about the neural guidance of eye movements is shaped by studying where we look in space and little is known about the role objects in our environment play in guiding eye movements. To address this gap, we recorded simultaneous intracranial electroencephalography with eye tracking in individuals with epilepsy during free viewing of scenes and during unscripted natural social interactions in the real world. In both paradigms, the category of object participants looked at next could be predicted based on their brain activity from prior to the onset of saccades to that object - even after controlling for the spatial location of the eye movement. Neural activity that enabled prediction of what a person was going to look at next arose from regions of parietal cortex traditionally associated with eye movement guidance, as well as category selective regions of ventral temporal cortex. The influence of para-foveal responses to the object of the next fixation was examined by comparing saccades traversing greater than 5 degrees of visual angle to smaller saccades, and prediction accuracy did not diminish for the larger saccades. Preliminary results show that next fixation prediction may involve theta frequency activity in category selective ventral visual cortex, putatively related to interactions with theta frequency activity in hippocampal regions involved in visual navigation. These results suggest that ventral temporal cortex may be involved in guiding what to look at next, with interactions between ventral temporal cortex, hippocampus, and parietal cortex then guiding where to look for that object.

Read the paper · More papers on PaperTik