EARBUD SIGNALS accurately estimated stroke volume in healthy adults, according to a feasibility study that used earbud microphones and deep learning to assess cardiovascular function outside traditional clinical environments.
The investigators reported strong agreement between predicted and reference stroke volume measurements, suggesting that widely available earbud technology could support future longitudinal cardiovascular monitoring.
Stroke volume, defined as the volume of blood ejected by the left ventricle during each contraction, is an important indicator of cardiovascular health.
Existing measurement approaches typically depend on specialised equipment and clinical expertise, limiting their accessibility for routine or continuous monitoring.
To address this challenge, researchers developed a deep learning artificial intelligence system that estimated stroke volume from cardiac signals captured by in-ear microphones.
The study included 23 healthy participants. In-ear audio was recorded using a custom earbud device and compared with stroke volume estimates obtained from a clinically validated monitoring system.
The system achieved a mean absolute error of 5.24 mL and a mean absolute percentage error of 6.83% when estimating average stroke volume.
Agreement between predicted and reference measurements was high. The researchers also found that 88% of the variation in measured stroke volume was explained by model predictions.
The percentage error was 11.05%, well below the 30% threshold considered acceptable for new stroke volume measurement devices.
Performance remained consistent before and after isometric exercise, with similar mean absolute errors observed across both conditions.
The researchers noted that demographic information alone and traditional feature-based approaches produced substantially poorer results, highlighting the value of the deep learning framework.
The findings suggest that in-ear cardiac signals contain sufficient physiological information to estimate stroke volume reliably, even in new users.
However, the study was limited to a relatively small cohort of healthy participants. The researchers proposed that future work should expand testing across broader populations, including patients with cardiovascular disease.
They also suggested that integrating the software into commercial earbuds could support affordable, continuous at-home cardiovascular monitoring and help identify changes in cardiac function over time.
Reference
Butkow KJ et al. Measuring cardiac stroke volume through in-ear audio sensing. Nat Commun. 2026;DOI:10.1038/s41467-026-75642-0
Featured image: Freepik on Adobe Stock
Each article is made available under the terms of the Creative Commons Attribution-Non Commercial 4.0 License.
