Back to Projects
class HeartClassifier:
def predict(self, audio):
feat = self.extract(audio)
return self.model(feat)
▶ Normal sinus rhythm
confidence: 94.7%
def predict(self, audio):
feat = self.extract(audio)
return self.model(feat)
▶ Normal sinus rhythm
confidence: 94.7%
Overview
HeartSound Intelligence is an AI-powered heart health monitoring device built on Raspberry Pi. It captures heart sounds through electronic stethoscope and performs real-time analysis using deep learning models.
Project Metrics
Solo Dev
Role
5 Modules
Scope
Edge Inference
Challenge
Provincial 3rd
Award
2 Copyrights
IP
Key Features
Deep Learning Classification
CNN-based heart sound classifier for multi-class anomaly detection
Real-time Signal Capture
USB audio interface with electronic stethoscope integration
Touchscreen GUI
5-inch PyQt5 multi-page interface with QStackedWidget routing
Trend Analysis
Historical data management and health trend visualization
Embedded Solution
Portable Raspberry Pi hardware with GPIO/USB control
Tech Architecture
Layered decoupled architecture with independent GUI and inference engine communication, hardware abstracted through standard interfaces.
PyQt5 QStackedWidget Python NumPy/SciPy TensorFlow PyTorch Raspberry Pi GPIO USB Audio
Frontend Backend AI / ML Infra
Interested in this project?
Let's discuss technical details or collaboration opportunities