Thyroid detection using Throat Chakra

A web-based application designed to analyze human voice features—symbolically linked to the throat chakra in historical medicine paradigms—and automatically detect early signs of Thyroid-related diseases using audio signal processing.

The project is structured into two main parts: a React + Vite Frontend for seamless user interaction and audio recording, and a robust Python Flask Backend for executing deep audio feature extraction utilizing Librosa.

🚀 Features

• In-Browser Audio Recording: Real-time microphone audio capture and direct file uploading.
• Audio Processing Engine: Computes fundamental vocal indicators used by vocal cord diagnostics:
  • RMS (Volume)
  • Pitch (F0)
  • Jitter
  • Hoarseness (HNR)
  • Vocal Fatigue / Spectral Centroid
• Google Sheets Integration: Automatically synchronizes extracted data into a Google Sheet instance for data persistence and audit trails.
• Microservice Architecture: Clean separation of frontend UI from backend prediction components allowing for easy ML model integrations.

---

🛠️ Tech Stack

Frontend Framework

• React.js (Vite Bootstrapper)
• Tailwind CSS (Styling & Responsiveness)
• Axios (API communication)

Backend Architecture

• Python 3
• Flask & Flask-CORS (REST API)
• Librosa & NumPy (Audio Signal Processing)
• gspread & oauth2client (Google API Interface)

---

🧩 Getting Started

Follow these steps to get the full stack up and running locally.

1. Backend Setup

The backend handles the /record and /predict routes, actively transforming user audio into readable feature coordinates.

# Navigate to the backend directory
cd "Main Project/Backend"

# Create a virtual environment and activate it 
# (For Windows: .\env\Scripts\activate)
python -m venv env
source env/bin/activate

# Install the Python dependencies
pip install -r requirements.txt

# Environment Setup
# Create a .env file locally (refer to .env.sample)
# Ensure you specify PORT, UPLOAD_FOLDER, and GOOGLE_CREDENTIALS_PATH
# Place the relevant GCP credentials JSON inside the appropriate folder.

# Run the Flask Server
python app.py

The backend server will start locally at http://127.0.0.1:8000.

2. Frontend Setup

The frontend serves as the main interactive page allowing dynamic data collection.

# Navigate to the frontend directory
cd "Main Project/Frontend"

# Install node dependencies
npm install

# Start the Vite development server
npm run dev

The application UI should now be accessible on http://localhost:5173. The proxy is pre-configured to communicate internally with the 8000 port of the backend.

---

🏗️ Structure Overview

📁 Main Project
├── 📁 Backend/
│   ├── app.py                     # Main Flask routing and request handling
│   ├── .env                       # Environment variables
│   ├── requirements.txt           # Python dependencies
│   ├── 📁 module/
│   │   ├── feature_extractor.py   # Librosa logic (RMS, Pitch, Jitter, etc.)
│   │   └── google_sheets.py       # Wrapper to insert rows safely to GCP
│   └── 📁 uploads/                # Secure temporary vault for audio uploads
└── 📁 Frontend/
    ├── vite.config.js             # React environment & setup (Server Proxy attached)
    ├── 📁 src/
    │   ├── 📁 components/         # Reusable structural components (like Recorder.jsx)
    │   ├── 📁 pages/              # Main routes (RecordingAudioPage.jsx)
    │   ├── 📁 styles/             # Global styles (index.css)
    │   ├── App.jsx                 # Main application component
    │   └── main.jsx                # Main entry point
    └── package.json               # Package Manifest

---

📈 Future Expansions

• Predictive ML Endpoint: The /predict backend endpoint currently prepares feature formatting; ready for substituting in a Pickled/Joblib Trained Classifier to return precise binary is_thyroid decisions.
• Patient Dashboard: Adding visualizations using Chart.js bridging existing historical data stored inside the Google Sheets.