Bence Bagi: Unsupervised discovery of discrete neural states in rats playing hide and seek

BCCN Berlin / Technische Universität Berlin

 

Abstract

Due to its inherent freedom and lack of experimental control, play behavior has received relatively little attention in neuroscience. In contrast to play, in classical neuroscience experiments neural activity is measured across many identical trials of animals performing relatively simple tasks with little ethological relevance, then analyzed by relating neural responses to pre-defined experimental parameters. In this thesis I present an alternative approach for exploratory data analysis on a single-trial level, applicable in more complex and naturalistic behavioral settings in which no two trials are identical. I analyze neural population activity in medial prefrontal cortex (mPFC) of rats engaging in the well-known game of hide-and-seek, and show that it is possible to discover what aspects of the task are reflected in the recorded activity, even with a limited number of simultaneously recorded cells. Using hidden Markov models I cluster population activity in mPFC into a set of neural states, each associated with a pattern in neural activity that reoccurs throughout the experiment. Despite variability in behavior from trial to trial, relating the inferred states to the events of the hide-and-seek game reveals neural states that consistently appear at the same phases of the trials. Furthermore, applying the segmentation inferred from neural data to the animals' behavior aids in demonstrating that mPFC activity reflects the actions and behavioral context of the animals, in a bottom-up exploratory way. Finally, I replicate the results in a second data set, and additionally show that neural activity in mPFC displays distinct sets of patterns during playing hide-and-seek and observing others play the game.

 

Additional Information

Master Thesis Defense

 

Organized by

Prof. Dr. Michael Brecht & Dr. Juan Ignacio Sanguinetti Scheck / Lisa Velenosi