Ethosoft
Prototype conceptHeart sounds · Signal processing

AI Digital Stethoscope

A digital-auscultation workflow for cleaning heart sounds, extracting acoustic features, and supporting normal-versus-murmur screening.

Project record

This page reports the maturity and evidence documented in Ethosoft's working archive. It does not imply peer review, regulatory clearance, or clinical validation.

Methods and evidence

Prototype conceptHeart sounds · Signal processing

AI Digital Stethoscope

A digital-auscultation workflow for cleaning heart sounds, extracting acoustic features, and supporting normal-versus-murmur screening.

Features
MFCC, RMS energy, zero-crossing rate, and normalized temporal representations.
Models considered
Temporal CNN/RNN approaches and classical SVM or random-forest baselines.
Interface
Mobile or web software paired with a Bluetooth-capable digital stethoscope.

Evidence boundary: the project file contains a system design but no reported patient-level evaluation.