In half of the cases tested, the AI’s judgment agreed with a cardiologist’s assessment.
As soon as your dog reaches middle age or senior years, it becomes more important to pay attention to their overall health, as well as early signs of any illness. Older dogs are more vulnerable to heart diseases, including the most common one: mitral valve disease. A heart murmur can be a sign of underlying condition but not all vets and nurses have excellent stethoscope skills to catch it. In the same way, there’s not enough time or specialists to screen every dog with a heart murmur. Which is why, an AI program that accurately detects heart murmurs could be the next big thing in the world of pet care.
In a new study published on Journal of Veterinary Internal Medicine on October 21, 2024, researchers reported about an AI program with 90% accuracy in detecting heart murmurs in canines. Not all murmurs are serious, though. But in half of the cases tested, the researchers found that the AI also agreed with an expert’s grade for the murmur’s seriousness.
According to Andrew McDonald, the lead investigator and a research associate in stethoscope acoustics at University of Cambridge, the grade of the heart murmur can determine the next steps that the vet and owner should take. By automating that process, he and his fellow researchers believe that the AI program could be an invaluable tool in giving dogs the best quality of life for as long as possible.
How the AI Algorithm Was Built
A database of heart sounds from dogs was needed to build the AI program. It wasn’t available during the course of the study but that didn’t stop the researchers. Anurag Agarwal, a professor of acoustics and biomedical technology at Cambridge and also one of the researchers, shared that they started by using a database of heart sounds from humans. The sounds were from 1,000 patients, gathered through a digital stethoscope. An AI program was then developed to check for heart murmurs and was adapted for dogs. Agarwal added that mammalian hearts are quite similar to each other, so they tend to go wrong in similar ways, too.
The AI program was put to a test at four vet clinics in the UK. It was able to acquire data from roughly 800 dogs that underwent routine heart exams. According to researcher and teaching professor of small animal cardiology at Cambridge, Dr. Jose Novo Matos, the more data they have, the more useful the algorithm would be. Mitral valve disease primarily affects smaller and adult dogs, but to further improve the program, they wanted to get data from dogs in different sizes, ages, and breeds.
By fine-tuning the AI program, it was able to grade heart murmurs and not just detect. Additionally, it can differentiate between murmurs caused by an advanced disease and those associated with a mild heart disease.
RELATED: Mark Cuban Declares Dogs Are Superior Problem Solvers to AI
Will AI Threaten Pet Care-Related Jobs?
When it comes to the grade of heart murmurs, the AI program tested offered the same judgement as the expert cardiologist. This happened for almost half of the cases tested, making the program a success. Naturally, the discussion went to AI as a threat to jobs. However, Novo Matos emphasized that it’s a tool that can make cardiologists better at what they do. It can also help vets and owners quickly identify dogs who need treatments the most.
Meanwhile, Agarwal stressed that they want to empower professionals in making the right decisions. McDonald also left a reminder that it’s up to the vets to detect a heart disease in dogs so it can be treated early.
RELATED: 10 Deadly Diseases Your Dog Can Catch From Other Dogs
From The Club
Sadly, our furry companions cannot c0mmunicate the pain or discomfort they have in the same way we can. That’s perhaps the most challenging part of making sure that our pups get the right medical attention and treatments. Hopefully, as artificial intelligence evolves in healthcare settings, it can offer helpful insights to professionals and shorten the time needed in identifying proper medications for different illnesses. AI holds great promise especially when used by experts with continuing medical education.
