UC Berkeley AI detects ECG signal predicting sudden cardiac death
University of California, Berkeley, researchers have trained AI to pick up a secret signal in standard ECG heart tests, something doctors usually can't see.
This breakthrough means the AI can now predict who might be at risk for sudden cardiac death, a condition that often strikes without warning, even if you pass regular checkups.
AI flags 7% sudden death risk
The team looked at more than 440,000 ECGs and found their AI could spot people with a 7% yearly risk of sudden death, way better than today's methods.
In fact, more than 86% of high-risk cases would have been missed by traditional screenings.
Lead researcher Ziad Obermeyer says this tech could help doctors figure out who needs extra testing or life-saving devices like defibrillators, making heart care smarter and more personalized.