A Convolutional Neural Network for Classification of EEG Signals from SSVEP Responses of Migraine

A CNN model developed as a feature extractor and classifier for SSVEP-based Brain-Computer Interface (BCI) for Migraine which may potentially be implemented in neurofeedback training or diagnostic protocol.

Note: This project is adapted from EEGNet architecture based on pytorch framework.

Workflow

— Data collection
— Data preprocessing
— Modeling
— Evaluation

You can read more about this project here!

Medium: link

Repository structure

— dataset
  • migraine (18 subjects)
  • control (18 subjects)
— preprocessed
  • feat (for model input)
  • label (class annotation)
— Jupyter notebook
— model

Requirements

MNE, Pytorch, Numpy, Pandas, Matplotlib, sklearn

Acknowledgement

This project is supported by AI Builders program