./TumourIdentifier <-- Folder containing first initial dataset from Kaggle ./TumourClassifier <-- Folder containing second initial dataset from Kaggle ./TumourDatasetFinal <-- Folder containing final combined dataset ./Output <-- Output results written from .ipynb files ./Images <-- Set of images used in our Latex report ./LogisticRegressionModel.ipynb <-- notebook to load data and run Logistic Regression Model ./ConvNetTumourIdentifier.ipynb <-- notebook to load data and run Convolutional Neural Network Model
- Install and import numpy, matplotlib, sklearn, openCV (cv2), imutils and tensorflow
- The files should be run in the order: LogisticRegressionModel.ipynb ConvNetTumourIdentifier.ipynb
- GPU is not required.
- Training takes ~2 hours.
- The report notebook saves files to the "./Output" directory and generates plots.
The TumourIdentifier data was downloaded from https://www.kaggle.com/navoneel/brain-mri-images-for-brain-tumor-detection The TumoutClassifier data was downloaded from https://www.kaggle.com/sartajbhuvaji/brain-tumor-classification-mri
For preprocessing and training the dataset we referenced the tutorial found at https://medium.com/@mohamedalihabib7/brain-tumor-detection-using-convolutional-neural-networks-30ccef6612b0