Detecting Brain Abnormalities using Self-Supervised Contrastive Learning

The repository contains the source code for detecting brain abnormalities using self-supervised contrastive learning. This work focuses on the detection and classification of different types of hemorrhages and was submitted as our undergraduate thesis.

Acknowledgement

We would like to thank and deep sense of gratitude to our guide Dr. CHITRA BABU, Professor, Department of Computer Science and Engineering, for her valuable advice and suggestions as well as her continued guidance, patience and support that helped us to shape and refine our work. We would like to thank Dr.U.S. SRINIVASAN, Senior Consultant Neurosurgeon, Sri Balaji hospital, Guindy, Chennai for providing the data and annotations for the project. We would like to extend our gratitude to him for helping us understand medical concepts.

Workflow of the project

1. Clone the repo

   git clone https://github.com/PrasannaKumaran/Detecting-Brain-Abnormalities-using-Self-Supervised-Contrastive-Learning.git

   For accounts that are SSH configured

    git clone git@github.com:PrasannaKumaran/Detecting-Brain-Abnormalities-using-Self-Supervised-Contrastive-Learning.git

2. Install pip

   python -m pip install --upgrade pip

3. Create and Activate Virtual Environment (Linux)

   python3 -m venv [environment-name]
   source [environment-name]/bin/activate

4. Install dependencies

   pip install -r requirements.txt

5. Important data folders list and ground truths can be found at data/

6. The data processing code is found at processsing/

7. Implementation details of the project can be found at implementation/

8. Read the cell description and run them one at a time

You can contribute to this work and make sure to cite our work!.

Team Members

Prasannakumaran D, Sideshwar J B, Sri Hari J