For a lot of people, atrial fibrillation shows up in flashes. A weird flutter, a fast thump, then nothing at all. The problem is that many of the tests we rely on work best when you already know to go looking. If the episode doesn’t happen during a clinic visit, it can slip by.
A research team led by Yan Chen and colleagues described a different approach in a study covered by Medical Xpress, using a radar-like sensor to read tiny mechanical motions from the beating heart without any physical contact. Instead of sticking electrodes to the skin, the system uses radio sensing and an AI model trained to recognize patterns linked to atrial fibrillation, borrowing what clinicians already know from ECG diagnosis and transferring that “knowledge” into the radar-based model.
In testing, the researchers evaluated the method using data from 6,258 outpatients, including 229 people with atrial fibrillation, collected during routine short ECG screenings. They also tried it in a real-life setting during sleep routines with 27 people, showing the idea isn’t limited to lab conditions. The hope is straightforward: if detection can happen quietly in the background, more people could catch atrial fibrillation earlier, before a stroke or other complications forces the issue.
This is still research, and it’s not a consumer device you can buy tomorrow. But it’s a positive signal in a hard area of medicine: making early detection easier for people who don’t feel symptoms, don’t wear monitors, or don’t know they’re at risk yet.
Source: https://medicalxpress.com/news/2025-05-contactless-radar-sensor-irregular-heart.html