This repository provides core codes and toolboxes for GCN model in the paper entitled "Using graph convolutional network to characterize individuals with major depressive disorder across multiple imaging sites" by Qin et al., 2022. Please see the paper for the detailed description.
Content includes source code for the implementation of graph convolutional network (GCN) on a large multi-site MDD fMRI dataset. All custom codes were tested on Linux Ubuntu 20.04 LTS PC.
- Python (>= 3.5)
- Scikit-Learn
- Pytorch
- Pytorch-geometric
- Scipy
- Numpy
- Pandas
Other toolboxes and codes used in our study for image preprocessing, harmonization, ancillary analysis and visualization are shown below: