The "Ocular Response Function" for encoding and decoding oculomotor related neural activity

Abstract

Oculomotor activity provides critical insights into cognition and health, with growing evidence demonstrating its involvement in various cognitive functions such as attention, memory, and sensory processing, and that it is a significant indicator of psychopathologies and neurological disorders. Despite its crucial importance across domains, the neural mechanisms supporting oculomotion have been underexplored, largely because eye movements are typically treated as artefacts to be removed from the neural signal. While useful for data cleaning, this approach risks discarding valuable information about oculomotor control, and there has been recent interest in modelling these signals instead to understand them. Using time-resolved regression methods with magnetoencephalography (MEG) and eye tracking during the resting state, we thus here sought to model Ocular Response Functions (ORFs), that characterise the neural signatures of distinct oculomotor events, specifically saccades, blinks, and pupil dilation. We demonstrate the relationships between ocular action and neural activity (encoding), revealing a range of sensory cortical, cerebellar and frontal signatures preceding and following such ocular events. We conversely show how we can reconstruct ocular events from brain activity, and further apply resting-state derived ORFs to a passive listening task - demonstrating how some neural markers interpreted as directly related to sensation may in principle indirectly result from oculomotor contributions to task-related neural processing. By providing an accessible framework for examining the interplay between eye movements and neural processes, we offer a range of insights with potential applications across cognitive and clinical neuroscience.