A deep learning pipeline for automated detection and classification of brain tumors from MRI scans using convolutional neural networks (CNNs). This project implements a complete medical image analysis workflow from preprocessing to model deployment.

This project implements an AI-powered system for brain tumor detection in MRI scans. The pipeline includes:

• Medical Image Preprocessing: Handling DICOM/MRI formats, normalization, and augmentation
• Deep Learning Model: Custom CNN architecture for tumor classification
• Model Training & Evaluation: Complete training pipeline with performance metrics
• Visualization Tools: Grad-CAM heatmaps for model interpretability

• ✅ Complete end-to-end deep learning pipeline
• ✅ Support for multiple CNN architectures (Custom CNN, ResNet, VGG)
• ✅ Data augmentation techniques for medical imaging
• ✅ Model interpretability with Grad-CAM visualizations
• ✅ Performance metrics and confusion matrix analysis
• ✅ Export trained models for deployment

The project utilizes a dataset of brain MRI images labeled as either "tumor" or "no tumor". The data is organized into training, validation, and test sets to ensure robust model evaluation.

This project is licensed under the MIT License. See the LICENSE file for more details.

• The project is inspired by the need for efficient and accurate medical diagnosis using AI.
• Special thanks to the creators of the dataset and the authors of the libraries used in this project.
• Dataset: https://www.kaggle.com/datasets/navoneel/brain-mri-images-for-brain-tumor-detection