Chest Cancer Classification

Project Overview

This project implements a deep learning solution for classifying chest CT scans to detect adenocarcinoma cancer. It uses a VGG16-based convolutional neural network to analyze medical images and classify them as either normal or showing signs of adenocarcinoma cancer.

Tech Stack

• Python 3.11
• TensorFlow 2.12.0
• Flask
• DVC (Data Version Control)
• MLflow
• Docker

Project Structure

The project follows a modular architecture with the following components:

• Data Ingestion: Downloads and extracts the dataset
• Base Model Preparation: Configures the VGG16 model for transfer learning
• Model Training: Trains the model with data augmentation
• Model Evaluation: Evaluates model performance and logs metrics

Features

• End-to-end ML pipeline with DVC for reproducibility
• Web interface for real-time predictions
• Docker containerization for easy deployment
• CI/CD pipeline using GitHub Actions
• AWS integration for model deployment

Installation & Setup

Prerequisites

• Python 3.11
• Docker (optional)

Local Setup

1. Clone the repository

git clone https://github.com/your-username/Chest-Cancer-Classification.git
cd Chest-Cancer-Classification

2. Create and activate a virtual environment

python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

Running the Application

Using Python

python app.py

Using Docker

docker build -t chest-cancer-classifier .
docker run -p 8080:8080 chest-cancer-classifier

Model Training Pipeline

To run the complete training pipeline:

python main.py

Or using DVC:

dvc repro

API Endpoints

• GET / - Web interface for uploading and classifying images
• POST /predict - API endpoint for image classification
• GET /train - Trigger model retraining

Project Configuration

Configuration parameters are stored in:

• params.yaml: Model hyperparameters
• config/config.yaml: Pipeline configuration

Deployment

The project includes a complete CI/CD workflow for AWS deployment:

1. Code is tested and linted
2. Docker image is built and pushed to Amazon ECR
3. Application is deployed to a self-hosted runner

This guy help me to land a AI Engineer job One day more