YUSTE LAB/COLUMBIA UNIVERSITYA program developed to filter spam email has been used to identify the behaviors of pond-dwelling hydras. In a study published last month (March 28) in eLife, researchers describe how they took an algorithm originally developed for language processing and adapted it to visually assess what the hydras did when they were presented with changing environmental conditions. The program can recognize predefined movements and pick out new ones on its own.
Hydras, which are closely related to anemones and jellies, do not have a central nervous system or brain. Instead, they possess a system of interconnected neurons called a nerve net, and the researchers hope they’ll be able to use the new technique to learn how the nerve net functions.
“People have used machine learning algorithms to partly analyze how a fruit fly flies, and how a worm crawls, but this is the first systematic description of an animal’s behavior,” says coauthor Rafael Yuste, a neuroscientist at Columbia University, in a statement. “Now that we can measure the entirety of hydra’s behavior in real-time, we can see if it can learn, and if so, how its neurons respond.”
S. Han et al., “Comprehensive machine learning analysis of Hydra behavior reveals a stable basal behavioral repertoire,” eLife, doi:10.7554/eLife.32605, 2018.