This project implements a Convolutional Neural Network (CNN) to classify brain MRI scans into four categories: Glioma Tumor, Meningioma Tumor, Pituitary Tumor, and No Tumor. The model is designed to automate the tumor detection process, aiding radiologists in early diagnosis and treatment planning.
- Source: Brain Tumor MRI Dataset on Kaggle
- Size: Approximately 7,000 MRI images
- Classes: Glioma Tumor, Meningioma Tumor, Pituitary Tumor, No Tumor
- Format: JPG images, organized into training and testing sets
📂 Brain_Tumor_Classification_CNN
│── Brain Tumors and Mental Health.docx # A brief report to decode the connection between brain tumor and mental health
│── README.md # Project documentation
│── brain-tumor-classification-using-cnn.ipynb # Jupyter notebook for training and evaluation
│── brain-tumor-classification-using-cnn.pdf # PDF export of the notebook
│── brain-tumor-classification-using-cnn.html # HTML export of the notebook
✔️ Data Preprocessing – Resized MRI images, applied data augmentation, and normalized pixel values.
✔️ Transfer Learning – Utilized ResNet18 for feature extraction.
✔️ Custom Classification Layer – Fine-tuned the final layers for multi-class classification.
✔️ Model Evaluation – Calculated accuracy, precision, recall, and F1-score.
✔️ Visualization – Included confusion matrices and loss curves for performance analysis.
- Overall Accuracy: 91% on test data
- High Precision and Recall for Pituitary and No Tumor classes
- Confusion Matrix to visualize model performance across all classes
- Deploy the model as a web application for real-time tumor detection
- Gandhar Ravindra Pansare (Indiana University, Bloomington)
- Guided by Professor Krista Li
This project is open-source under the MIT License.