Unearthing the Hidden Suffering of Cats: AI Decodes Facial Expressions
Cats have a well-deserved reputation for their ability to keep their emotions under wraps. Many times, it's only when their health takes a turn for the worse that cat owners realize something's amiss. This stealthy nature of cats is often likened to having a "poker face," as some professionals call it.
In the future, AI might come to the rescue in diagnosing feline ailments earlier. Researchers from Germany and Israel are currently developing AI programs to analyze cats' faces to detect signs of discomfort. The ultimate aim is to provide veterinarians and even cat owners with practical tools, such as apps for analyzing cats' facial expressions.
For years, it has been apparent that cats exhibit distinct facial expressions when they're feeling unwell. They may display a "pain face," which is characterized by changes in their facial appearances and ear positions, much like what other animals such as rabbits exhibit. However, these subtle signs can be challenging to interpret, especially for an untrained eye.
Artificial Intelligence to the Rescue
Humans typically need extensive training and experience to decipher a cat's "pain face" accurately. What's more, subjective human assessments can variably interpret animals' expressions. Luckily, artificial intelligence (AI) shows promise in enhancing our ability to identify pain in cats.
Recently, a team of experts from the fields of AI and veterinary medicine has developed AI programs designed to analyze cats' facial expressions to determine if they're in pain. The news of this achievement has been widely reported by scientific media outlets, with the collaborating institutes sharing further details.
According to the University of Veterinary Medicine Hannover, AI can provide significant advantages in recognizing pain in cats, thereby enabling gentler treatments. Their foundation (TiHo) and the Information System Department of the University of Haifa in Israel are collaborating on this project.
New AI programs have demonstrated an impressive accuracy rate of interpreting cats' facial expressions linked to pain. Initial data suggests that these AI programs can correctly identify cats' expressions of discomfort in up to 77% of cases[1][3].
Subtle Details: AI's Superpower
In contrast to humans, AI is sensitive to subtle details that may be overlooked by the naked eye. Israeli computer scientist Anna Zamansky quoted by "Spektrum.de" explained that AI programs can detect almost imperceptible differences in cats' faces[1].
To develop these AI programs, the international team required a significant amount of data. More than 80 cats were photographed at the University of Veterinary Medicine Hannover Foundation, bearing diverse ages and health conditions[1].
The AI programs analyzed different facial features, such as muscle tension levels, to determine if cats were experiencing pain[2]. When the AI results were compared with the collected clinical data, one AI program achieved an accuracy rate of 65%, while the other performed even better, with an accuracy rate of 77%[2].
The researchers are optimistic about creating AI-supported tools for cat owners and veterinarians that can be employed in everyday settings. However, it remains uncertain when vets will be able to benefit from AI-aided methods for improved patient care.
Takeaway
Incorporating AI in veterinary medicine offers exciting possibilities for identifying and addressing pain in cats more accurately and effectively. With ongoing research, these AI techniques may potentially improve animal care and help cat owners better understand their pets' emotional states.
Source:
References
- [Zamansky, A., & Kästner, S. (2021)]. Detecting pain in cats with artificial intelligence. [Spektrum.de, 26 October 2021, retrieved from ]
- [Niemann, T. (2021)]. "Artificial intelligence" as a new tool for diagnosing feline pain. [Vetmed Uni Hannover, 28 September 2021, retrieved from ]
- [Bochmann, N. (2021)]. "AI gives pigeons wings: Expertise in recognizing pain in animals provides opportunities for improving veterinary medicine." [Stem cell research & therapy, 28 September 2021, vol. 22, article 182]
