Mentor: Shaun Sze-Xian Lim
Evaluating the behavior of mice is essential to evaluating the effects of various biochemical and physical manipulations and shapes the way for elucidating new drugs and therapies that can potentially translate into the clinical setting. The 3D-assisted Neural Network for Computational Ethology (DANNCE) algorithm was developed to more quantitatively describe mice behavior by making predictions based on a machine learning algorithm. However, currently, there is not a standardized device, or arena, in which to place these mice and implement the algorithm. Our lab developed an arena, which by tuning mirror angles, can capture 5 perspectives of a mice enclosed in a rectangular box with one high-resolution camera. By using a large sample of mice in this arena, we hope to establish the efficacy of our design and standardize mice behavior observation. In our experiment, we alter the phase of the circadian rhythm of the mice, their dosage of caffeine, and observe the effect of gender on different mice behavior. We are currently working on completing the data set collection and hope to see preliminary predictions that are consistent with our expected observations.
