Brain Tumor Classifier

About The Project

A comprehensive deep learning solution for automated brain tumor classification using MRI scans. This project leverages a state-of-the-art EfficientNet architecture to accurately detect and classify three types of brain tumors: glioma, meningioma, and pituitary tumors.

The system is deployed as a production-ready web application that allows medical professionals to upload MRI scans and receive instant predictions with visualization of the model's decision process using GradCAM technology.

Built with modern MLOps practices, the pipeline includes containerized deployment on AWS cloud infrastructure, managed through infrastructure-as-code principles.

Key features:

• Real-time tumor classification with confidence scores
• Advanced visualization of model attention using GradCAM
• Scalable cloud deployment using AWS EC2 and ECR
• Infrastructure automation with Terraform
• Interactive web interface built with Streamlit

Stack

• Deep Learning: PyTorch, torchvision
• Cloud Services: AWS (EC2, ECR)
• Infrastructure: Terraform
• Containerization: Docker
• Web Interface: Streamlit
• Development: Python, Git

Directory Structure

  root_directory/
  │
  ├── app.py
  ├── Dockerfile
  ├── data/
  ├── models/
  ├── notebooks/
  ├── terraform/
  ├── utils/
  ├── README.md
  └── requirements.txt

Streamlit Web App deployment via EC2 and ECR

Installation

ECR local setup

1. Install AWS CLI if not installed

   brew install awscli
   

2. Configure AWS credentials

   aws configure
   

3. Create ECR repository

   aws ecr create-repository --repository-name <repository-name>
   

4. Get ECR login credentials and login

   aws ecr get-login-password --region <aws-region> | docker login --username AWS --password-stdin <aws-account-id>.dkr.ecr.<aws-region>.amazonaws.com
   

5. Tag image for ECR

   docker tag <docker-image-name>:<tag> <aws-account-id>.dkr.ecr.<aws-region>.amazonaws.com/<docker-image-name>:<tag>
   

6. Push image to ECR

   docker push <aws-account-id>.dkr.ecr.<aws-region>.amazonaws.com/<docker-image-name>:<tag>
   

EC2 setup

1. Provision infrastructure with Terraform

   terraform -chdir=terraform apply

2. Download the .pem file and set permissions:

   chmod 400 <your-key-file.pem>

3. Connect to the EC2 via SSH

   ssh -i </path/to/key-file.pem> ec2-user@<instance-public-ip>

4. Install Docker

   sudo yum install docker -y
   sudo service docker start
   sudo usermod -a -G docker ec2-user
   newgrp docker

5. Install AWS CLI

   curl "https://awscli.amazonaws.com/awscli-exe-linux-x86_64.zip" -o "awscliv2.zip"
   unzip awscliv2.zip
   sudo ./aws/install

6. Configure AWS region

   aws configure
   

7. Log in to ECR

   aws ecr get-login-password --region <aws-region> | docker login --username AWS --password-stdin <aws-account-id>.dkr.ecr.<aws-region>.amazonaws.com
   

8. Pull container

   docker pull <aws-account-id>.dkr.ecr.<aws-region>.amazonaws.com/<docker-image-name>:<tag>
   

9. Run container

   docker run -d -p 8501:8501 <aws-account-id>.dkr.ecr.us-east-1.amazonaws.com/<docker-image-name>:<tag>
   

10. Access the Stramlit app using the EC2 public IP address:

    http://<ec2-public-ip>:8501
    

Useful Commands:

• To destory the infrastructure provisioned

  terraform -chdir=terraform destroy
  

Contributing

• Headstarter.co
• Matteo Pilotto