search

elgazzarr / DAST-GCN

Official Pytorch Implementation of the paper Dynamic Adaptive Spatio-temporal Graph Convolution for fMRI Modelling
MIT License
16 stars 4 forks source link

readme

Dyanmic Adaptive Spatio-temporal GCN for fMRI modelling (DAST-GCN)

This is the original pytorch implementation of DAST-GCN in the paper Dynamic Adaptive Spatio-temporal Graph Convolution for fMRI Modelling

Requirements

Data Preparation

See dataframe samples at csvfiles

Infered Graph Visulizations

Infered graph after training for both tasks on ukbb. 'The graph was restricted to static configuration ('1 adaptive grah structure for all blocks') for more robust visulizations'