TL;DR
Get everyday helpers delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
A study presented at the American Heart Association meeting found that an AI model using smartwatch ECGs could identify reduced left ventricular ejection fraction, but results varied between Apple and Samsung watches and across repeated readings from the same person. Researchers said the findings do not establish how well the approach would work as a screening tool in broader, real-world populations.
A Korean study presented at the American Heart Association annual meeting found that an AI model analyzing smartwatch ECGs could identify reduced heart pumping function, but its readings were not consistently aligned across Apple Watch and Samsung Galaxy Watch devices or across repeated home recordings by the same person. The findings point to limits that researchers say need to be addressed before smartwatch ECGs can be used reliably to screen for left ventricular systolic dysfunction, or LVSD.
The prospective study involved 2,211 patients at Incheon Sejong Hospital who were undergoing echocardiography for clinical reasons. Participants wore both an Apple Watch and a Galaxy Watch for ECG recordings. The researchers applied an AI model trained to detect LVSD and used a shared cutoff, previously set using a separate Galaxy Watch tuning group. The study defined LVSD as an ejection fraction of 40% or less, measured within 14 days of an echocardiogram.
The model’s area under the curve was 0.907 for Apple Watch and 0.895 for Galaxy Watch. Reported sensitivity was 69.2% and 53.4%, respectively, while specificity was 88.9% and 93.4%. Despite those performance measures, 9% of participants had discordant results at the shared cutoff: 158 tested positive only on Apple Watch, while 42 tested positive only on Galaxy Watch. Hak Seung Lee, the study presenter, said Apple Watch scores tended to be higher.
A separate group of 146 people recorded ECGs at home using one smartwatch, most often a Galaxy Watch. In that group, 49.3% had both positive and negative readings during the same week at the study cutoff. Attempts to make the results more consistent, including averaging readings over a week, had limited success. A stricter rule reduced false positives but also missed more cases of LVSD, according to the report.
Why Device-to-Device Differences Matter
Smartwatch ECG screening could make it easier to identify people with reduced heart function, including those who have not yet been diagnosed. But different answers from different devices, and mixed results from repeated readings on one device, could complicate decisions about who should receive follow-up testing. A positive result may lead to further evaluation, while a negative reading could be falsely reassuring if the model misses LVSD.
The findings also highlight a health-system challenge: screening can help only if people with concerning results are connected to confirmatory testing and appropriate care, while false positives do not overwhelm clinical services. Jessica Golbus, a University of Michigan cardiologist who participated in the meeting press conference, said the goal is to identify LVSD while managing the burden of false alarms. The study did not establish whether use of the technology improves diagnosis, treatment, or patient outcomes.
Products Worth Considering
Samsung Galaxy Tab S7 11-inch Android Tablet 128GB Wi‑Fi with S Pen Mystic Black
Samsung Galaxy Tab A11+ 8GB RAM, 256GB Storage, Gray
Samsung Galaxy Tab A9+ 11-Inch Android Tablet, 64GB, Graphite
As an affiliate, we earn on qualifying purchases.
How the Study Tested Smartwatch ECGs
Interest in this approach followed earlier research. A 2022 study reported good accuracy for an AI model using a smartwatch’s single-lead ECG to detect LVSD. The current study tested a model called AiTiALVSD-1L, which Lee said was pretrained on nearly 3 million unlabeled ECGs and fine-tuned using about 360,000 clinical lead I ECGs.
The study set a cutoff of 23.5 in a separate Galaxy Watch tuning set to produce 98% specificity, then applied that cutoff to both watch brands. The main device comparison group was not a general screening population: 6.6% had known LVSD and 31% had known heart failure. In the separate home-recording group, 33.6% had LVSD and 55.5% had known heart failure. These characteristics matter when interpreting how well the results might apply to people without a prior diagnosis.
“A single negative home recording is not reassuring on its own, and mixed results should prompt further evaluation.”
— Hak Seung Lee, MD, study presenter
Limits of the Current Evidence
The study was conducted at one hospital and did not test screening performance in asymptomatic people without known heart failure. Its participants were already undergoing clinically indicated echocardiography, so the results may not carry over to people using watches at home without a specific reason for cardiac testing.
It is also unclear how performance varies across age, race, ethnicity, and levels of ejection fraction. The study did not show that smartwatch-based detection changes treatment or health outcomes. The reported accuracy measures and device comparisons do not by themselves establish that either watch can diagnose LVSD; confirmation and clinical evaluation remain outside what these results demonstrate.
Validation Before Wider Screening
Researchers said further work should test device-specific models and cutoffs in larger, more varied populations, including people without known heart failure. Studies will also need to determine how clinicians should respond to positive, negative, or mixed readings and whether the approach improves patient outcomes without creating excessive follow-up testing.
For now, the report describes an investigational screening approach rather than a validated replacement for clinical assessment. The next milestone is evidence showing how smartwatch readings perform beyond this single-center study and whether acting on them leads to earlier treatment or better outcomes.
Key Questions
What did the study find about smartwatch detection of LVSD?
The AI model showed measurable ability to distinguish LVSD in the study sample, but results differed between Apple and Samsung watches. In a separate home-recording group, 49.3% had both positive and negative readings during the same week.
Does a negative smartwatch ECG rule out reduced heart function?
No. The study presenter said a single negative home recording was not reassuring on its own. The study does not establish smartwatch ECGs as a way to rule out LVSD.
What did the researchers mean by LVSD?
The study defined LVSD as a left ventricular ejection fraction of 40% or less, measured within 14 days of an echocardiogram.
Can these findings be applied to everyone who owns a smartwatch?
That is not known. The study took place at one hospital among people referred for echocardiography, and researchers did not assess screening in asymptomatic people without heart failure.
What evidence is still needed?
Further research needs to test performance across broader populations and watch devices, and determine whether screening leads to better clinical outcomes while keeping false-positive follow-up manageable.
Source: medpagetoday.com
