Brain-Tumor-Classification

Check out the project

https://brain-tumor-classification-syw4urqyf2jue5aftr43eg.streamlit.app/

Running Locally

1. Clone the repository:

   git clone https://github.com/aaespino238/Brain-Tumor-Classification.git
   cd Brain-Tumor-Classification

2. Install Dependencies:

   pip install requirements.txt

3. Set Up Environment: Create a .env file in the root directory and configure one or more of the following variables

   GEMINI_API_KEY=your-gemini-api-key
   OPENAI_API_KEY=your-openai-api-key
   GROQ_API_KEY=your-groq-api-key

   Support for additional models can be added by creating a proper provider file in utils/aisuite/providers/ and adding it to the MODELS list with the appropriate name in utils/llm_utils.py.
   • Refer to https://github.com/andrewyng/aisuite for more information.
4. Run the Application Locally:

   streamlit run app.py
   

Additional Information

• All model training related code is located is src

• Model training code was heavily inspired by https://github.com/oarriaga/face_classification

• Dataset: https://www.kaggle.com/datasets/masoudnickparvar/brain-tumor-mri-dataset

  • Dataset contains a number of duplicates, duplicate removal code is located in src/utils/remove_duplicates.ipynb

• Trained models can be downloaded using the download_models function located in utils/models.py.

  • Code for download_models was referenced from https://medium.com/@bridog314/deploying-large-ml-model-files-to-streamlit-using-google-drive-0818b0d416c9.