It’s using AI.
A recent study reveals that artificial intelligence (AI) can effectively identify heart murmurs in dogs, achieving an impressive accuracy rate of 90%, comparable to that of experienced cardiologists. Researchers found that in over half of the tested cases, the AI’s evaluations aligned perfectly with those of a cardiologist regarding the severity of the murmurs.
Heart murmurs are significant indicators of heart valve disease, which is the most prevalent cardiac condition among adult dogs. According to the study, approximately one in every thirty dogs examined by a veterinarian presents with a heart murmur, a condition that is more common in older and smaller breeds. Early identification of this issue is vital, as prompt treatment can significantly prolong the lives of affected dogs.
Lead researcher Andrew McDonald, a research associate specializing in stethoscope acoustics at the University of Cambridge, emphasized the importance of early detection. He noted that while heart disease is a major concern for humans, it poses an even greater challenge for dogs, particularly small breeds, as they are likely to develop heart issues with age. Since dogs cannot communicate their health problems like humans, it falls to veterinarians to recognize these conditions in a timely manner.
The research team initially utilized a database of heart sounds from about 1,000 human patients collected through a digital stethoscope. They adapted this information to create an AI program specifically for detecting murmurs in dogs, acknowledging the lack of existing canine heart sound databases. By leveraging the similarities in mammalian heart function, they aimed to develop a robust tool for veterinary use.
The AI was tested on nearly 800 dogs receiving standard heart examinations at various veterinary clinics across the U.K. The researchers specifically focused on gathering diverse data from dogs of different breeds and ages to enhance the AI’s effectiveness. The algorithm was designed to detect and assess the severity of murmurs based on digital recordings, distinguishing between those caused by mild conditions and those linked to more advanced heart disease. With the AI agreeing with expert cardiologists in over half of the assessments, it provides a promising supplementary resource for veterinarians to improve the quality of care for canine patients.
The study highlights that, unlike humans who primarily rely on surgery for heart valve disease treatment, there are various medications available for dogs. Understanding when to initiate treatment is crucial for ensuring the best possible quality of life for dogs with heart conditions. The researchers see AI not as a threat to veterinary jobs, but rather as a valuable tool that can enhance diagnostic capabilities and streamline the process of identifying dogs in need of treatment.
Discussion about this post