Rodent Arena Tracker (RAT): A Machine Vision Rodent Tracking Camera and Closed Loop Control System

Abstract

Video tracking is an essential tool in rodent research. Here, we demonstrate a machine vision rodent tracking camera based on a low-cost, open-source, machine vision camera, the OpenMV Cam M7. We call our device the Rodent Arena Tracker (RAT), and it is a pocket-sized machine vision-based position tracker. The RAT does not require a tethered computer to operate and costs about $120 per device to build. These features make the RAT scalable to large installations and accessible to research institutions and educational settings where budgets may be limited. The RAT processes incoming video in real-time at 15Hz and saves X and Y positional information to an onboard microSD card. The RAT also provides a programmable multi-function input/output pin that can be used for controlling other equipment, transmitting tracking information in real time, or receiving data from other devices. Finally, the RAT includes a real-time clock (RTC) for accurate time stamping of data files. Real-time image processing averts the need to save video, greatly reducing storage, data handling, and communication requirements. To demonstrate the capabilities of the RAT, we performed three validation studies: 1) a 4-day experiment measuring circadian activity patterns; 2) logging of mouse positional information alongside status information from a pellet dispensing device; and 3) control of an optogenetic stimulation system for a realtime place preference (RTPP) brain stimulation reinforcement study. Our design files, build instructions, and code for the RAT implementation are open source and freely available online to facilitate dissemination and further development of the RAT.

Using Head-Mounted Ethanol Sensors to Monitor Olfactory Information and Determine Behavioral Changes Associated with Ethanol-Plume Contact during Mouse Odor-Guided Navigation

“…In the future, we plan to combine cameras with actuators in a closed-loop system that would follow the animal and move the wires accordingly. This approach is possible using relatively low-cost components (Fee and Leonardo, 2001;Krynitsky et al, 2020). The small footprint of these sensors can be leveraged to allow us to record the olfactory information arriving at the two nares simultaneously.…”

An open-source device for measuring food intake and operant behavior in rodent home-cages

“…We provide example programs for running rodents on multiple common behavioral paradigms, including free feeding, time-restricted free-feeding, fixed-ratio (FR1), and progressive ratio (PR) operant tasks, and have written a FED3 Arduino library to ease development of custom programs. FED3 also has an auxiliary I2C port for hardware customization and a programmable BNC output connection that allows synchronization with external equipment for aligning behavioral events with fiber-photometry recordings (London et al, 2018;Mazzone et al, 2020), electrophysiological recordings (London et al, 2018), or other equipment such as video tracking systems (Krynitsky et al, 2020;Li et al, 2019). This BNC output can report task events or control external hardware with delays of less than 200µs, which compares favorably with computers that process information through Universal Serial Bus (USB) which can introduce delays of several milliseconds.…”

An Infrared Touch System for Automatic Behavior Monitoring

“…Video recording, the most common method in behavioral data acquisition, records the entire experimental process. Large amounts of information can be extracted from videos, such as animal trajectories, posture, the facial expressions, and even whisker movements [2,[22][23][24][25][26][27][28]. Based on the rapid development of computer vision and machine learning, numerous algorithms have been developed to extract more and more information from videos, as well as increasing processing speed.…”