Purchase this article with an account.
Jayden Brennan, Brent Young, Ning Tian; Virtual reality behavior testing for visual perception. Invest. Ophthalmol. Vis. Sci. 2016;57(12):2764.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Virtual reality (VR) behavioral tests have frequently been used to study navigation and memory in mice, but it is rarely used to study visual perception. We developed a VR behavioral task that can test recognition of various visual cues by mice.
Mice were placed in custom harnesses and trained to be able to run on a spherical treadmill coupled to a computer mouse in order to navigate in a virtual world. Mice were conditioned to two different types of visual cues in order to test visual perception. One group was trained to arrive at a target that consisted of black and white bars moving downwards. Other targets consisted of bar moving up, down, left or right as background targets (directional). Another group of mice was trained to arrive at a green target with the other background targets being blue, light grey, and dark grey (color). During training mice were initially introduced to running on the spherical treadmill and receiving their drink reward. The next step consisted of introducing the mice to the virtual arena and guiding them to the correct target. Lastly mice were allowed to navigate on their own while data tracing their path was recorded on the computer. Multiple measures such as number of rewards per session, average distance to reward, average velocity, and average time to reward were recorded as the mice navigated in the virtual arena.
Of those average distance to reward, and average time to reward were found to have an influence on the variance in the amount of rewards obtained for both groups of mice. A linear regression model showed that increased average velocity was strongly indicative of an increase in the amount of rewards for directionally trained mice, but was not for color trained mice (P < 0.0001 for directional, P = 0.96363 for color). Mice removed from any form of training for a month were found to only require 5-7 additional days of booster training before reward performance was back to baseline compared to 13 days of initial training.
Our data concludes that mice are able to be trained using a VR rig for testing of visual perception. Additionally, it appears that different visual stimuli produce different behavioral and training responses. More detailed analysis needs to be performed on color and directional mice groups to determine the cause of these differences.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.
This PDF is available to Subscribers Only