Harmful Brain Activity Classification (HBAC)

This model aims to classify seizures and other patterns of harmful brain activity in critically ill patients. We'll be using data provided by Harvard Medical School, as part of an ongoing Kaggle competition.

Important Links

Timesheet	Slack channel	Project report

Video/demo/GIF

Table of Contents

1. Demo

2. Installation

3. Reproducing this project

4. Guidance

1. Example demo

For the demonstration, we created a web interface that identifies neurological disorders and displays the output to users upon receiving data input. We used Python flask, which stores our HTML files and serves as the location for our Python code. After using flask run to start a flask server, we put the single data file as a CSV file and upload the data, the analysis result shows at the bottom ['GPD,' 'GRDA,' 'LPD,' 'LRDA,' 'Other,' 'Seizure'].

What to find where

repository
├── calculations                 ## Testing correctness for feature values
├── data                         ## EEG sample datasets 
├── images                       ## Images for our doc and demo
├── src                          ## Source code for our package itself
│   └── feature_extraction/
│   └── preprocessing/
│   └── utils/
│   └── visualize/
│
├── static/css                   ## Styling for our HTML demo
├── templates                    ## HTML templates for our demo          
├── README.md                    ## You are here
├── requirements.yml             ## Dependencies
├── hms.ipynb                    ## Jupyter notebook
├── hms.py                       ## Python file for our Jupyter notebook

2. Installation

To clone the repository and install required dependencies please run:

git clone git@github.com:sfu-cmpt340/2024_1_project_11.git
cd 2024_1_project_11
conda env create -n 340-project-11 -f requirements.yml
conda activate 340-project-11

3. Reproduction

Our python script contains code that automatically retrieves 1000 signals from our dataset.

To reproduce our model results, run the following:

conda activate 340-project-11
python hms.py

Alternatively, you can run the Jupyter notebook hms.ipynb.

4. Guidance

• Use git
  • Do NOT use history re-editing (rebase)
  • Commit messages should be informative:
    • No: 'this should fix it', 'bump' commit messages
    • Yes: 'Resolve invalid API call in updating X'
  • Do NOT include IDE folders (.idea), or hidden files. Update your .gitignore where needed.
  • Do NOT use the repository to upload data
• Use VSCode or a similarly powerful IDE
• Use Copilot for free
• Sign up for GitHub Education