🧠 Brain Tumor MRI Image Classification (Deep Learning + Streamlit)

This project is a deep learning–based medical imaging application that classifies brain MRI images into different tumor categories using Custom CNN and Transfer Learning models.
A Streamlit web app is provided for real-time tumor prediction from uploaded MRI images.

---

🎯 Problem Statement

To build an AI-powered system that can automatically classify brain MRI images into tumor types such as:

• Glioma
• Meningioma
• Pituitary
• No Tumor

This helps radiologists and doctors with faster diagnosis, early detection, and decision support.

---

🏥 Real-Time Business Use Cases

• AI-Assisted Medical Diagnosis
• Early Detection & Patient Triage
• Clinical Research & Trials
• Second-Opinion AI for Telemedicine

---

🚀 Features

• MRI image classification using Deep Learning
• Custom CNN model from scratch
• Transfer Learning with MobileNet
• Image preprocessing & normalization
• Model evaluation & confidence score
• Streamlit-based interactive web app
• Google Drive–based model loading (for deployment)
• Clean & simple UI for doctors and users

---

🧾 How to Run this Project

• git clone https://github.com/keshavjatt/Brain-Tumor.git
• cd Brain-Tumor
• pip install -r requirements.txt
• python train_custom_cnn.py
• python train_transfer.py
• python evaluate.py
• python -m streamlit run app.py