AI can now detect if cats are in pain by scanning feline faces

An app has been trained to detect pain in cats faces by scanning thousands of pictures of felines. The developers of the artificial intelligence app say it is more than 95% accurate.

Sky News, Thursday 13 June 2024

https://news.sky.com/story/ai-can-now-detect-if-cats-are-in-pain-by-scanning-feline-faces-13152363

Which cat is in pain? The AI is trained to spot that the one on the left is not in pain whereas the ginger cat on the right is. Pic: Careology

Artificial intelligence (AI) can now detect whether your cat is in pain by looking at its face, using an app built by vets and developers in Japan.

The AI app was trained using 6,000 pictures of cats, and developers say it is more than 95% accurate.

The head of the Animal Medical Centre at Nihon University worked on the app. He said vets like him can tell to a certain degree whether an animal is in pain or not, but it's harder for owners.

"More than 70% of elderly cats have arthritis or pain, but only 2% of them actually go to a hospital," said Professor Edamura.

"So rather than a final diagnosis, we use [the app] as a tool to make owners aware of whether the situation is normal or not."

More than 230,000 people have used the app since it launched in Japan last year.

The cat photos used to train the AI app were designed to enable it to tell when a cat is feeling pain. Pic: Reuters

Mayumi Kitakata lives in Tokyo with her 14-year-old cat Chi. She uses the app to read his face each day.

"He is at an age where more and more diseases are going to appear," said Ms Kitakata.

The head of Nihon University Animal Medical Center, Professor Kazuya Edamura, gives a lecture to students on the diagnosis of pain in cats. Pic: Reuters

Pets are very popular in the country, with the Japan Pet Food Association estimating there were almost 16 million pet cats and dogs in the country last year, more than the number of children under 15.

The app, called CatsMe!, was developed by Careology and researchers at Nihon University.