arnnv/xray-classification

MediSync Backend

This is the backend service for MediSync, providing API endpoints for medical image analysis and report generation using AI models.

Prerequisites

• —Python 3.10 or higher
• —pip (Python package manager)
• —Docker (optional, for containerized deployment)

Project Structure

backend/
├── api.py              # Main API endpoints
├── app.py             # FastAPI application setup
├── predict.py         # Prediction logic
├── langgraph_agent.py # LangGraph agent implementation
├── requirements.txt   # Python dependencies
├── Dockerfile        # Docker configuration
├── weights/          # Directory for model weights
└── .env             # Environment variables

Setup Instructions

1. Local Development Setup

1. 1.Create and activate a virtual environment (recommended):

   python -m venv venv
   # On Windows
   .\venv\Scripts\activate
   # On Unix or MacOS
   source venv/bin/activate

1. 1.Install dependencies:

   pip install -r requirements.txt

1. 1.Set up environment variables:

• —Create a .env file in the backend directory
• —Add required environment variables:

     GOOGLE_API_KEY=your_google_api_key_here

1. 1.Download model weights:

• —Create a weights directory in the backend folder
• —Download the required model weights and place them in the weights directory

1. 1.Run the development server:

   python app.py

The server will start on http://localhost:7860

2. Docker Setup

1. 1.Build the Docker image:

   docker build -t medisync-backend .

1. 1.Run the container:

   docker run -p 7860:7860 --env-file .env medisync-backend

API Endpoints

The backend provides the following main endpoints:

• —POST /predict: Upload and analyze medical images
• —POST /generate-report: Generate medical reports based on analysis
• —Additional endpoints are documented in api.py

Dependencies

Key dependencies include:

• —TensorFlow (2.10.x)
• —LangChain & LangGraph for AI processing
• —FastAPI for API endpoints
• —Google Generative AI for report generation
• —Additional dependencies listed in requirements.txt

Environment Variables

Required environment variables:

• —GOOGLE_API_KEY: API key for Google's Generative AI services

Development Guidelines

1. 1.Code Structure:

• —api.py: Contains all API endpoint definitions
• —predict.py: Handles image processing and model predictions
• —langgraph_agent.py: Implements the LangGraph agent for report generation

1. 1.Adding New Features:
2. 2.Add new endpoints in api.py
3. 3.Implement corresponding logic in appropriate modules
4. 4.Update requirements.txt if new dependencies are added

Troubleshooting

Common issues and solutions:

1. 1.Model weights not found:

• —Ensure model weights are properly placed in the weights directory
• —Check file permissions

1. 1.API key issues:

• —Verify .env file exists and contains valid API key
• —Check environment variable loading in application

1. 1.Dependencies issues:
2. 2.Try removing the virtual environment and recreating it
3. 3.Update pip: pip install --upgrade pip
4. 4.Install dependencies one by one if bulk installation fails