Probing visual sensitivity and attention in mice using reverse correlation

Abstract

Visual attention is a fundamental cognitive operation that allows the brain to evoke behaviors based on the most important stimulus features. Although mouse models offer immense potential to gain a circuit-level understanding of this phenomenon, links between visual attention and behavioral decisions in mice are not well understood. Here, we describe a new behavioral task for mice that addresses this limitation. We trained mice to detect weak vertical bars in a background of checkerboard noise while audiovisual cues manipulated their spatial attention. We then modified a reverse correlation method from human studies to link behavioral decisions to stimulus locations and features. We show that mice attended to stimulus locations just rostral of their optical axis, which was highly sensitive for vertically oriented stimulus energy whose spatial frequency matched those of the weak vertical bars. We found that the tuning of sensitivity to orientation and spatial frequency grew stronger during training, was multiplicatively scaled with attention, and approached that of an ideal observer. These results provide a new task to measure spatial- and feature-based attention in mice which can be leveraged with new recording methods to uncover attentional circuits.