The system consists of a box fitted with two cameras, one at the top and one facing the mouse, similar to a photo booth.
Keystone-SDA
An AI monitoring system developed by the Swiss federal technology institute ETH Zurich uses algorithms to recognise changes in the facial expressions of rodents caused by stress or pain.
+ Get the most important news from Switzerland in your inbox
The new Grimace system standardises the monitoring of laboratory animals, the ETH Zurich said in a statement on Monday. It uses machine vision and machine learning to assess the facial expressions and postures of mice. According to the researchers, observation is immediate, gentle and objective.
The system consists of a box fitted with two cameras, one at the top and one facing the mouse, similar to a photo booth for identity photos. The two cameras are used to detect subtle signs of pain and discomfort that are often apparent in the facial expressions of rodents: squinted eyes, bulging noses and cheeks, changes in the position of the ears or the orientation of the whiskers.
More
More
How AI is used in diagnostics
An algorithm then evaluates the mouse’s facial expressions in real time. This system makes it possible to quickly and accurately assess whether the animals are suffering and whether additional analgesics need to be administered.
Computer versus naked eye
Monitoring pain and well-being is essential for animal experimentation. But until now, assessment has been subjective and often inadequate. Scientists observe the animals from the edge of their cages and compare their facial expressions with photographs, which often proves difficult. Human observation can also cause the animals additional stress.
More
More
How Swiss AI recognises wild animals – and helps biologists worldwide
Researchers at the ETH Zurich tested the Grimace system to assess the well-being of mice before and after brain surgery. The computer assessments were consistent with those of human experts. Their main advantage was that they were consistent and standardised, whereas the judgement of pain levels differed from one expert to another. These results were published in the specialist magazine Lab Animal.
Open source kit
The entire system, including the software, was developed by staff at the ETH Zurich’s 3R hub, which is responsible for animal experimentation, based on the three principles of “replace, reduce and refine”. It is now being shared with the world as an open source kit.
Automated approaches to facial recognition already existed, but what was missing was a fully standardised end-to-end system, said Oliver Sturman, director of the 3R hub. Grimace is already attracting a great deal of international interest: the researchers plan to develop this technology, while the questions of patenting and marketing remain open.
Translated from French with AI/gw
We select the most relevant news for an international audience and use automatic translation tools to translate them into English. A journalist then reviews the translation for clarity and accuracy before publication.
Providing you with automatically translated news gives us the time to write more in-depth articles. The news stories we select have been written and carefully fact-checked by an external editorial team from news agencies such as Bloomberg or Keystone.
If you have any questions about how we work, write to us at english@swissinfo.ch.
