AI has just revealed that mice have a lot of facial expressions that help express their feelings - thereby giving us fresh evidence of how emotional responses form in the human brain.
Scientists at the Max Planck Institute for Neurobiology in Germany have discovered this by recording the faces of mice in the laboratory when they are stimulated by various factors, such as sweetness, or electric shock. The researchers then used machine learning algorithms to analyze how the rodents' faces changed when they experienced different sensations.
The software shows that the facial muscles of the mouse will move slightly when their emotions change . Rats experiencing pain will shrink their ears and puff out their cheeks, while happy mice will push their jaws and ears forward, and pull their noses toward their mouths.
The AI also revealed that mice 'expressions will vary, depending on how they feel before being stimulated. For example, when a thirsty mouse is given sugary water, its face will show more joy than an already thirsty mouse.
The researchers then investigated which brain cells triggered those reactions using a technique called photolysis, which uses light to excite different neurons. They found that when targeted at neurons capable of activating various emotions, mice displayed corresponding facial expressions.
The facial muscles of the mice will move slightly as their emotions change.
The idea of facial expressions that can reveal how the brain processes emotions is not new. In 1872, Charles Darwin realized that humans and animals communicate emotionally through similar expressions. Recent advances in AI have opened the door for us to test his hypothesis.
" I am very excited about how we humans have emotional states that we experience as sensations, " says neuroscientist Nadine Gogolla, who led a three-year study. . " I want to see if we can know how those states form in the brain through animal studies ."
Gogolla's team trained the software to recognize mouse responses by providing it with pictures of the mouse's facial expressions with related emotion notes. They then showed the system unnoticed images and asked it to decode the emotions in the picture.
AI can identify emotions with 90% accuracy - according to an article in Science.
The researchers believe that the information they gathered about how neurons trigger the facial reactions of rats will help us understand how the human brain handles anxiety disorders and depression. And they will probably prove that Darwin was always right!