The effect of locomotion on early visual contrast processing in humans

Most of our knowledge about vision comes from experiments in which stimuli are presented to immobile human subjects or animals. In the case of human subjects, movement during psychophysical, electrophysiological or neuroimaging experiments is considered to be a source of noise to be eliminated. Animals used in visual neuroscience experiments are typically restrained and, in many cases, anaesthetized. In reality however, vision is often used to guide the motion of awake, ambulating organisms. Recent work in mice has shown that locomotion elevates visual neuronal response amplitudes (Erisken et al., 2014; Fu et al., 2014; Lee et al., 2014; Mineault et al., 2016; Niell and Stryker, 2010) and reduces long-range gain control (Ayaz et al., 2013). Here we use both psychophysics and steady-state electrophysiology to ask whether similar effects of locomotion on early visual processing can be measured in humans. Our psychophysical results show that brisk walking has little effect on subjects9 ability to detect briefly-presented contrast changes and that co-oriented flankers are, if anything, more effective masks when subjects are walking. Our electrophysiological data were consistent with the psychophysics, indicating no increase in stimulus-driven neuronal responses whilst walking and no reduction in surround suppression. In summary we find evidence that early contrast processing is altered by locomotion in humans but in a manner that differs from that reported in mice. The effects of locomotion on very low-level visual processing may differ on a species-by-species basis and may reflect important differences in the levels of arousal associated with locomotion. SIGNIFICANCE STATEMENT Mice are the current model of choice for studying low-level visual processing. Recent studies have shown that mouse visual cortex is modulated by behavioural state: V1 neurons in locomoting mice tend to be more sensitive and less influenced by long-range gain control. Here we test these effects in humans by measuring psychophysical detection thresholds and EEG responses while subjects walk on a treadmill. We find no evidence of increased contrast sensitivity or reduced surround suppression in walking humans. Our data show that fundamental measurements of early visual processing differ between humans and mice and have important implications for recent work on the link between arousal, behaviour and vision in these two species.