Mayo Clinic heart ECG AI is being rolled out to more doctors

In short: An AI tool from Mayo Clinic that can spot hard-to-see heart risk signals in ECGs is being made available to more clinicians through licensing deals.

What happened

Mayo Clinic researchers built AI models that read electrocardiograms, also called ECGs (a simple test that records the heart’s electrical signals). The models were trained on millions of ECGs and linked health outcomes, so they learned patterns that can be too subtle for people to notice.

The program is now reaching more hospitals and clinics through commercial partners. Mayo says one FDA-cleared version for a standard 12-lead ECG can help detect low left-ventricular ejection fraction, which is a weak heart pump. Mayo has licensed this 12-lead tool to Anumana.

Mayo has also licensed a single-lead version, which can work with handheld devices, to Eko Health. Because the AI is being added inside systems that hospitals already use for ECGs or digital stethoscopes, some clinicians can access the results without buying separate software themselves. That is why coverage sometimes describes it as “free” to doctors, even though health systems and vendors still handle contracts and costs.

Why it matters

ECGs are quick, common, and relatively cheap. If AI can reliably flag problems like a weak heart pump, hidden atrial fibrillation (an irregular heartbeat that can come and go), hypertrophic cardiomyopathy (abnormally thick heart muscle), or cardiac amyloidosis (protein buildup in the heart), it could help patients get follow-up tests sooner. It is best viewed like an extra set of eyes, not a final diagnosis, so clinicians still need to confirm results and consider the full medical picture.

Source: NYTimes