Brain Tumor Classification app

This is my major project, which attempts to solve a medical challenge of tumor detection achieving a high-accuracy using transfer learning from a pre-trained model. This app uses a EfficientNetB0 pretrained model to classify images.

Kaggle notebook

The kaggle notebook for the same is https://www.kaggle.com/code/architjee/brain-tumor-classification-from-mri

The kaggle notbook also contains step-by-step process of training and using and finally testing the model.

Then I used the trianed model, exported and converted it into a streamlit web-app, currently hosted at https://share.streamlit.io/architjee/braintumorclassifier/main.py

About the application

This app takes an image ( Brain MRI ) input and classifies it into one of the following 4 categories:

1. No Tumor
2. Meningioma Tumor
3. Glioma Tumor
4. Pituitary Tumor

Screenshot

Installation

To run the application type in the terminal/powershell

pip3 install -r requirements.txt

Usage

Followed by

streamlit run main.py

Hosted at

Also hosted at https://share.streamlit.io/architjee/braintumorclassifier/main.py Probably would have to turn off your adblocker to use it.