AI reduces the use of laboratory mice in research
Published: Thursday, Nov 14th 2024, 16:20
Retour au fil d'actualité
Researchers at ETH Zurich have developed a new, AI-supported method that analyzes the behaviour of mice in the laboratory more efficiently. As a result, biomedical research requires fewer laboratory mice.
The method uses automatic behavioral analysis by means of image recognition and artificial intelligence (AI). Mice are filmed and the video recordings are automatically analyzed, writes ETH Zurich in a press release issued on Thursday.
The analysis of animal behavior often meant days of laborious manual work, writes de ETH. In recent years, leading research laboratories worldwide have switched to powerful automated behavioral analysis. But the mountains of data still accumulated.
And: the more extensive the data and the finer the differences in behavior to be recorded, the greater the risk of misinterpretation, according to the ETH.
Stress resistance tested
The ETH researchers' new method now makes it possible to obtain meaningful results even with a smaller group of animals and to identify subtle behavioral differences between the animals, the report continues. The new approach also helps to reduce the number of test animals. And the standardization of tests can be increased.
Using the new method, the ETH team has already been able to find out, among other things, how mice react to stress and certain drugs in animal experiments. And that acute and chronic stress change the behavior of mice differently.
©Keystone/SDA
French 
English 
German 
Spanish 
Arabic