eMurmur Heart AI

eMurmur Heart AI (2.2) is a software decision support system that analyzes recorded heart sounds to identify S1 and S2 heart sounds, detect abnormal heart murmurs or their absence (including innocent murmurs), and determine the loudness of individual heart sound components. It is intended for use by healthcare providers in settings where auscultation can be performed properly, as a decision-support aid rather than a sole diagnostic tool for pediatric and adult patients.

eMurmur Heart AI (2.2) is hosted on a secure, encrypted backend server, requires digital heart sound recordings from supported digital stethoscopes and patient age as input, and interfaces with the eMurmur Transformer software accessory for multi-stethoscope compatibility. The new version adds a feature determining average loudness of heart sound components and their ratios, with enhanced cybersecurity including private networking that exposes only the backend API to the internet.

Not stated in this summary.

eMurmur Heart AI (2.2) is substantially equivalent to its predicate (eMurmur Heart AI 1.0) because both share the same intended use (analyzing heart sounds to aid clinician evaluation), similar indications, and identical user population and principles of operation. Performance data show non-inferior sensitivity (90.0% vs 86.7%), specificity (90.0% vs 88.3%), and accuracy (90.0% vs 87.5%) compared to the predicate, with the new loudness-analysis feature validated and raising no new safety or effectiveness questions.