kierenschmidthsg/neurosight-brain-tumor-models

NeuroSight Brain Tumor Classification Models

Pre-trained TensorFlow/Keras models for classifying brain MRI scans into four categories:

• Glioma
• Meningioma
• Pituitary tumor
• No tumor

Model Details

• Architecture: Custom CNN with 4 convolutional layers
• Framework: TensorFlow/Keras
• Input: 192×192 grayscale MRI images
• Test Accuracy: ~91%
• Test Recall: ~90%

Usage

Download and extract the models:

wget https://huggingface.co/kierenschmidthsg/neurosight-brain-tumor-models/resolve/main/neurosight-models.tar.gz
tar -xzf neurosight-models.tar.gz

This creates two directories:

• brain_tumor_cnn_improved/ - Main classification model
• brain_tumor_cnn_keras/ - Model for Grad-CAM visualizations

Training Dataset

Trained on the Brain Tumor MRI Dataset from Kaggle.

Repository

Full project code: GitHub Repository

License

CC0 1.0 Universal (same as training dataset)

Disclaimer

This is an educational project and should not be used for medical diagnosis.