EA-GNAS

PyTorch Source code for "Optimizing Graph Neural Network Architectures for Schizophrenia Spectrum Disorder Prediction Using Evolutionary Algorithms".

You can also find other open-sourced biomedical signal analysis projects in my academic page. ☺️ ☺️ ☺️

Data Preparation

• HC/SCH data needs to be downloaded and placed in Data/HC_SCH/
• The user needs to obtain data usage rights from the webpage https://bicr-resource.atr.jp/srpbsfc/

How to run this project

This project contains one main file as follows:

Searching the proper structure of the graph neural network

python main_IOAGraph.py

• Here, you can get the proper model structure （Search flag = True）.
  • Switch different intelligient optimization algorithms.
• Also, you can train, test, and explain the proper model structure （explain flag = True）.

If our work is helpful to you, please Star it and kindly Cite our paper as:

@article{WANG2024108419,
title = {Optimizing graph neural network architectures for schizophrenia spectrum disorder prediction using evolutionary algorithms},
journal = {Computer Methods and Programs in Biomedicine},
volume = {257},
pages = {108419},
year = {2024},
issn = {0169-2607},
doi = {https://doi.org/10.1016/j.cmpb.2024.108419},
url = {https://www.sciencedirect.com/science/article/pii/S0169260724004127},
author = {Shurun Wang and Hao Tang and Ryutaro Himeno and Jordi Solé-Casals and Cesar F. Caiafa and Shuning Han and Shigeki Aoki and Zhe Sun}
}