An Alternative Low-Cost Solution for Tracking Laboratory Animals

Abstract

Objectives: Monitoring animal behavior under various conditions can provide important information on their neuropsychological status, including learning, memory, and cognitive activity. Several commercial tracking systems are available, but they may be too expensive for low-budget projects. This study developed an alternative solution for automated animal tracking in behavioral experiments.

Methods: The proposed system was designed to analyze a set of images sampled from a recorded video file in chronological order. The instantaneous location of the animal in each image frame was defined automatically, using a feature-extraction algorithm. Distances traveled were calculated using the coordinates of the successive instantaneous locations. The algorithm was tested using two arenas: the Morris water maze and open field test. The calculated measures were compared with those obtained manually. The internal consistency of the dataset was checked using Cronbach's alpha. The accuracy of the results was evaluated using the paired samples t-test and Pearson correlation, with the level of statistical significance set at p<0.01.

Results: A statistical comparison of the distances traveled, which were derived from the coordinates of successive locations, did not differ significantly between the manual and automatic methods (r=0.954 and p=0.792 for the Morris water maze; r=0.996 and p=0.024 for the open field test).

Conclusions: These results suggest that the algorithm is reliable and valid for estimating coordinates and may serve as a high-resolution tool for animal behavior experiments. We intend to make this software freely available to interested readers and to open feedback channels for further development.