Title:
A Highly-Scalable Automated Rodent Training Platform

Thumbnail Image
Author(s)
Ali, Yahia H.
Authors
Advisor(s)
Pandarinath, Chethan
Advisor(s)
Editor(s)
Associated Organization(s)
Supplementary to
Abstract
Background: Rodent training is a necessary but time-consuming process that often requires the development of computer-based training systems to automate the process of administering a food or water reward in return for the rodent performing the desired behavioral task. To increase the throughput of these systems, they need to be scaled up to simultaneously train more rodents, but current scalable automated systems are incompatible with the graphical programming software used to develop behavioral tasks. New Method: Here, we present a novel scalable rodent training platform that allows researchers to scale behavioral tasks developed in graphical programming environments such as Simulink and LabVIEW. This system communicates with training cages over a network, so it is compatible with all internet-enabled devices regardless of the software or operating system they are running. Results: The system is validated by training a cohort of four previously-untrained rats over a period of three weeks in a fully automated fashion. In ten sessions or less, all of the rats learned to extend to touch a knob in return for a food reward. Comparison with Existing Method(s): The use of this training system to control two training cages running Simulink software and automatically adjust the training parameters according to each rat’s performance represents an advance over previous single-cage training systems. Conclusions: In this study, the ability to train rats on a novel forelimb perturbation task demonstrates the complex behavioral tasks that can now be studied in a scalable, automated fashion while maintaining compatibility with the graphical programming tools currently in use.
Sponsor
Date Issued
2019-05
Extent
Resource Type
Text
Resource Subtype
Undergraduate Thesis
Rights Statement
Rights URI