Abstract
Mice have emerged as a useful animal model for studying aspects of the visual system, but have not yet been shown to exhibit visual selective attention. In fact, the discrepancies between the visual and cognitive abilities of primates and mice raise the possibility that mice might not have selective visual attention, at least not in the forms well-known in primates. Here we tested for selective visual attention in mice, using three behavioral paradigms adapted from primate studies of attention. We trained head-fixed mice running on a wheel to detect a threshold-level change in a visual stimulus with distractors, and provided visual spatial cues that indicated the likely location of the relevant visual change. Because our mice reported their detection by licking a central spout, and the visual change occurred in either the right or left visual field, we were able to measure lateralized effects on perceptual choice without confounds from biases in motor responses. We found spatially specific changes in perceptual sensitivity, criterion and reaction times. In a Posner-style cueing task, a spatial cue indicated the probable location of the relevant visual event, and we found that valid cues increased response accuracy and shortened reaction times. In a cue versus no-cue task, an informative spatial cue was provided on half the trials, and we found that the spatial cue again increased response accuracy and shortened reaction times, and lowered detection thresholds measured from psychometric curves. In a filter task, the spatial cue indicated the location of the relevant visual event, and we found that mice could be trained to ignore irrelevant but otherwise identical visual events at uncued locations. Together, these results demonstrate that mice exhibit visual selective attention, paving the way to use visual paradigms in mice to study the genetic and neuronal circuit mechanisms of selective attention.
Meeting abstract presented at VSS 2018