Quantitative Behavioral Technology

In the wild, mice use their vision to serve many adaptive tasks such as navigating, hiding from predators, foraging, and hunting. Many such adaptive tasks can be useful to neuroscientists trying to study the underlying neuronal processes that guide these behaviors. Recently, it has become increasingly possible to reproduce these behaviors in a lab setting, using virtual reality, to engage our neural processes of interest in repetitive, quantitative frameworks. In SLAB, we have a novel virtual foraging task designed to interrogate visual object recognition, or the process allowing us to identify objects despite various visual transformations, wherein the mice treadmill through a 3D space to find a target object for a food reward. 

My project involves analyzing the behavioral data generated by mice performing this task to measure performance and other behavioral parameters as they learn, in order to optimize the training protocol. In addition, I’m assisting in improving the current behavioral setup, and when covid allows, I will assist in training the mice to complete the virtual task myself.

The results from this project will benefit neuroscientists trying to understand the neural processes supporting visually-guided behaviors. More broadly, understanding the neural algorithms underlying visual processing could lead to greater advancement of machine learning and artificial intelligence.