Detecting pulmonary hypertension and low ejection fraction from a digital stethoscope and three-lead ECG — catching conditions that rarely announce themselves (Eko Devices, NCT07087613)
A prospective observational study of whether heart sounds and three-lead electrocardiograms recorded with a digital stethoscope can help detect pulmonary hypertension and a left ventricular ejection fraction of 40 percent or less. Echocardiography or right heart catheterization serves as the reference standard, with up to 3,850 participants planned.
Trial overview (primary data)
- StatusRecruiting
- ConditionsHypertension, Pulmonary, Heart Failure With Reduced Ejection Fraction
- InterventionsDEVICE: Eko CORE 500 Digital Stethoscope
- SponsorEko Devices, Inc.
- Target enrollment3,850 participants
- Period2025-06-15 〜 2027-05-31
Key points
- A prospective observational study of detecting pulmonary hypertension and low ejection fraction from heart sounds and a three-lead ECG recorded with a digital stethoscope.
- Participants complete a single session of about 20 minutes, with an echocardiogram or right heart catheterization within seven days serving as the reference standard.
- The record notes pulmonary hypertension carries significant risk if undiagnosed and low ejection fraction can go undetected for lack of symptoms.
- Up to 3,850 participants may be enrolled across multiple sites so that roughly 3,500 complete.
- Of the 803 medical-AI trials this site holds as of 2026-08-31, 5 touch stethoscopes, against 33 for chatbots, 26 for documentation and 22 for wearables.
- Of the 803 trials held as of 2026-09-01, 33 concern chatbots while only 5 touch a stethoscope.
1Putting AI on an old method
The stethoscope has been in use for more than two centuries and still sits at the entrance to an examination. It is cheap, present everywhere, and takes a few minutes with the patient lying down. What can be heard through it, though, has always depended heavily on the experience of whoever is listening. What this study attempts is to record that sound digitally and let a machine judge.
Rather than adding a new test, it raises the information yield of a step everyone already goes through.
2Conditions that go unfound because nothing shows
Both targets named in the record are easy to miss. Pulmonary hypertension carries significant risk if it progresses undiagnosed, and low ejection fraction can remain undetected because patients have no symptoms or only nonspecific ones. Echocardiography and right heart catheterization will reveal them, but those are tests performed once there is a symptom or a suspicion.
Without something to prompt suspicion, the test is never reached. Catching it at the stethoscope stage could change that order.
3Stethoscopes appear in 5 of 803 trials
Sorting the bodies of the 803 medical-AI clinical trials this site holds as of 2026-08-31 by keyword, only 5 touch stethoscopes. Chatbots appear in 33, documentation or scribing in 26, and wearables in 22. Clinical application of AI gathers around reading images, writing text and holding conversations, leaving listening to sound comparatively thin.
At up to 3,850 participants this study is among the larger ones this site holds as of 2026-08-31. How accurate the detection proved is not part of this registry record.
4Where clinical AI is concentrating
Classifying the text of the 803 medical AI trials this site holds as of 2026-09-01 by keyword shows that where AI is being used is far from evenly spread.
Clinical AI gathers around reading images, writing text and holding conversations, and listening to sound remains comparatively thin. What this study attempts is recording, digitally, the sound of a stethoscope used for more than two centuries, and having a machine judge it. It does not add a new examination; it raises the information content of a step everyone already goes through.
Why it matters
Raising what can be gathered without adding a test lowers the cost of finding things earlier. Layering AI onto a step everyone already undergoes widens coverage with no extra visit and no extra fee. Debate about medical AI gathers around images and text, and this distribution shows sound remaining comparatively untouched.
FAQ
Why a stethoscope?
Why are these conditions hard to find?
Has detection been shown to work?
Sources (primary)
Source: ClinicalTrials.gov (U.S. NIH/NLM, public domain). This site does not provide medical advice. Verify the latest and exact details with the official source. This site is not endorsed or certified by the NIH/NLM.
- ClinicalTrials.gov (study record, original)
- NCT ID: NCT07087613