Human clinicians outperform AI stethoscope in diagnosing pet heart problems
Fourth-year veterinary students matched—and experienced clinicians outperformed—an AI-enabled digital stethoscope in diagnosing heart murmurs and arrhythmias in cats and dogs, according to a new study from North Carolina State University. The study emphasizes the importance of human expertise when using AI-assisted technology, particularly with instruments designed for "universal" usage. The…
A new study from North Carolina State University has found that fourth-year veterinary students outperformed an AI-enabled digital stethoscope in diagnosing heart problems in cats and dogs. The stethoscope, which is designed to help listen to heartbeats, was found to have limitations when it comes to detecting murmurs and arrhythmias in animals.
The research team compared the stethoscope's findings with those of board-certified veterinary cardiologists, cardiology residents, and fourth-year veterinary students. In dogs, the human experts found that 70% had a murmur and 44% had an arrhythmia. The AI stethoscope correctly identified 87% of the murmurs in dogs but failed to identify any arrhythmias.
For cats, the human experts found that 43% had a murmur and only 2% had an arrhythmia. The stethoscope was able to detect murmurs in only 2 out of 51 cats, missing 20 of the true murmurs. The study highlights the importance of human expertise when using AI-assisted technology, especially for devices designed for universal usage.
Written by urgent.news from Phys.org's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
