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
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.