❤️ AI Spots Heart Valve Disease More Accurately Than Doctors

Heart valve disorders are often described as a "silent epidemic." They affect more than half of people over 65, yet can remain undetected for years because symptoms may be mild or nonexistent. Researchers at the University of Cambridge have developed an AI algorithm that analyzes heart sounds recorded by a digital stethoscope to identify serious valve problems.

🎧 The team recorded heart sounds from 1,800 older adults, then confirmed diagnoses using echocardiography—the most accurate method for diagnosing valve disease.

After training on the audio recordings and test results, the algorithm learned to pick up subtle acoustic patterns that even experienced doctors can't reliably hear.

🔍 The system detected 98% of severe aortic stenosis cases, a life-threatening condition that leads to death within two years in about 80% of untreated patients. Overall, the algorithm achieved 72% sensitivity in identifying valve disease and recommended unnecessary follow-up tests in only 18% of cases. Doctors who listened to the same recordings detected 62% of cases and recommended unnecessary tests twice as often.

It works quickly—the AI requires only up to 15 seconds of heart sound recording. That could mean faster screening, lower costs, and fewer patients waiting weeks for an echocardiogram.

@hiaimediaen