Deep Learning for Algorithmic Detection of Pulmonary Hypertension Using a Combined Digital Stethoscope and Three-lead Electrocardiogram

This is a prospective, observational, multi-site study to determine whether synchronized heart-sound (phonocardiogram; PCG) and three-lead electrocardiogram (ECG) recordings (entered as separate interventions in PRS, though collected together in practice) collected with the FDA (U.S. Food and Drug Administration)-cleared Eko CORE 500 digital stethoscope can assist in screening for pulmonary hypertension (PH). Adults (≥18 years) referred for clinically indicated transthoracic echocardiography (TTE) and/or right heart catheterization (RHC) will complete a single study visit (~20 minutes). During the visit, trained staff will obtain at least four 15-second CORE 500 recordings from standard auscultation locations (aortic, pulmonic, tricuspid, and mitral).

The clinical TTE (and RHC, if performed) completed within ±7 days of the recordings will provide reference labels for the presence and severity of PH. If available, de-identified results from a clinical 12-lead ECG performed within 30 days of the TTE may also be abstracted for comparison. The primary objective is to develop and validate a software algorithm that detects PH from PCG+ECG signals; secondary objectives include assessing severity stratification and overall diagnostic performance (e.g., sensitivity/specificity, area under the receiver operating characteristic curve [AUC], positive predictive value [PPV], negative predictive value [NPV]), including subgroup analyses to evaluate generalizability. These recordings are investigational data acquisitions for algorithm development and detection research only; they are not diagnostic procedures and will not be used for clinical decision-making.

The study plans to enroll up to ~1,513 participants to obtain approximately 1,375 evaluable datasets (about 1,250 with TTE and 125 with RHC). No clinical decisions will be based on investigational algorithm outputs, and participation does not alter standard care. Key eligibility includes adults able to consent who have a clinical TTE or RHC within the protocol window; exclusions include current hospitalization and limited TTE studies. All study data are de-identified using site-maintained key-coded identification codes (IDs). PCG/ECG recordings are transmitted to a HIPAA (Health Insurance Portability and Accountability Act)-compliant, secure server; de-identified demographics, clinical data (e.g., TTE Doppler measures including TR [tricuspid regurgitation] velocity and estimated pulmonary artery systolic pressure, RHC hemodynamics), and relevant reports/images may be abstracted/uploaded per site procedures.

This minimal-risk study has IRB (Institutional Review Board) approval. No independent data monitoring committee is appointed; safety oversight is provided by site investigators with adverse event reporting per IRB policy. There is no cost to participants; modest compensation (≤$50, site-specific) may be provided. The CORE 500 hardware is FDA-cleared; however, the software algorithm evaluated in this study is investigational (not FDA-cleared). The aim is regulatory readiness for potential future submission without committing to a specific regulatory filing as part of this study.