A team of researchers from the University of Michigan has created a new open-source software tool that utilizes artificial intelligence to make animal behavior analysis easier for researchers across the life sciences. The new software, LabGym, is designed to identify, categorize, and count specific behaviors across various animal model systems.
Animal behavior analysis is crucial in research to understand how certain drugs affect organisms, map out the communication of brain circuits, and study the development and functions of the nervous system, among other reasons. For example, researchers in the lab of U-M faculty member Bing Ye study movements and behaviors in Drosophila melanogaster, commonly known as fruit flies, which offer insights into human health and disease as they share many genes with humans.
Analyzing animal behavior is a function of the brain that provides essential information on how the brain works and changes in response to disease, according to Yujia Hu, a neuroscientist in Ye’s lab at the U-M Life Sciences Institute and lead author of the study published in Cell Reports Methods.
While several software programs exist to automatically quantify animal behaviors, they have limitations as they are based on pre-set definitions of behavior. Therefore, Hu and his colleagues designed LabGym to more closely replicate the human cognition process and be more user-friendly for biologists who may not have expertise in coding.
Using LabGym, researchers can input examples of the behavior they want to analyze and teach the software what it should count. The program then uses deep learning to improve its ability to recognize and quantify the behavior. The software can also track multiple animals simultaneously, making it more efficient than traditional methods that require manual counting.
One new development in LabGym is the use of both video data and a pattern image to improve the program’s reliability. The pattern image shows the animal’s movement pattern by merging outlines of the animal’s position at different time points, making it easier for the program to identify behaviors.
LabGym’s species flexibility is also a significant advantage as it can be adapted to any species without requiring programming skills or high-powered computing. The program was originally designed for Drosophila but is not restricted to any one species, making it highly versatile.
U-M pharmacologist Carrie Ferrario, who studies the neural mechanisms that contribute to addiction and obesity, offered to help Ye and his team test and refine the program in the rodent model system she works with after hearing a presentation about the program’s early development. Ferrario and her team rely heavily on hand-scoring to complete necessary observations of drug-induced behaviors in rats, which is subjective and extremely time-consuming. LabGym offers a solution to this existing problem, and Ferrario sees its potential to be useful in almost limitless conditions to analyze animal behavior.
The team plans to continue refining the program to improve its performance under more complex conditions, such as observing animals in nature. The research was supported by the National Institutes of Health.
In addition to Ye, Hu, and Ferrario, the study authors include Alexander Maitland, Rita Ionides, Anjesh Ghimire, Brendon Watson, Kenichi Iwasaki, Hope White, and Yitao Xi of the University of Michigan, and Jie Zhou of Northern Illinois University.