This project was done in the Computational Neuroscience Lab (CNS Lab) at Stanford. The main goal of the project is to predict sex and age from resting state functional MRI (f-MRI) scans using a Spatio-Temporal Graph Convolutional Network. The accepted paper can be found at https://arxiv.org/abs/2003.10613
This is a simplified version of the ST-GCN module: https://github.com/yysijie/st-gcn
The previous version of this module implementation for f-MRI scans can be found at https://github.com/ericksiavichay/cs230-final-project
DATA USAGE
For HCP, the the data "Resting State fMRI 1 Preprocessed" was downloaded from https://db.humanconnectome.org/. The GIFTI images were then parcellated using https://balsa.wustl.edu/QnXj
Here are the final processed BOLD signal files: https://drive.google.com/file/d/1c4UYvp089KwqAllAHdHMHGjq8y5x882D/view.
If you need the NCANDA data, you should directly contact last author Dr. Kilian Pohl ([email protected]) for data usage.