🌬️ Lung Auscultation Assessment System (LAAS)

The Lung Auscultation Assessment System (LAAS) is a project I originally developed in 2022.
It combines augmented reality guidance with machine learning models to help users perform lung sound auscultation more accurately and consistently.

---

Table of Contents

• Features
• Tech Stack
• Screenshots
• How to Run
• Next Steps
• Credits

---

Features

• AR Guidance: Step-by-step overlay to position the stethoscope correctly.
• Machine Learning Integration: CNN-based model to analyze lung sounds.
• Accessibility Focus: Low-cost, user-friendly design for preventive healthcare.

---

Tech Stack

• Frontend: Swift / SwiftUI (AR integration)
• ML Model: PyTorch
• AR: Apple ARKit overlays
• Backend/Deployment: AWS (prototype for model serving)

---

Product Screenshots

Anterior View

Posterior View

---

How to Run

Prerequisites

• macOS with Xcode 14+
• iOS device (iPhone/iPad) with iOS 15+
• Python 3.9+
• PyTorch installed
• AWS CLI (optional, for model serving)

Steps

# Clone the repo
git clone https://github.com/YOUR_USERNAME/LAAS.git
cd LAAS

iOS App

1. Navigate to /ios-app/
2. Open LAAS.xcodeproj or LAAS.xcworkspace in Xcode
3. Run on a physical iOS device (ARKit not supported in simulator)

ML Model

cd ml-model
pip install -r requirements.txt
python predict.py --input sample_sound.wav

(Optional) AWS Deployment

• Upload model to S3
• Deploy with Flask/FastAPI
• Update app config with API endpoint

---

Next Steps

• Retrain ML model with larger, more diverse datasets
• Enhance AR overlay precision
• Deploy as a testable iOS app prototype

---

Credits

Created by Sophie Lin