This is the Pytorch implementation of KnowMDD in the paper: [KnowMDD: Knowledge-guided Cross Contrastive Learning for MDD Diagnosis].
- pytorch
- pytorch-geometric
- pandas
- nilearn
To run the code
- download the datasets REST-meta-MDD Project from DIRECT Consortium, which you can download from: https://www.scidb.cn/en/detail?dataSetId=cbeb3c7124bf47a6af7b3236a3aaf3a8.
- find the
ROISignals_FunImgARCWFfolder and theREST-meta-MDD-PhenotypicData_WithHAMDSubItem_V4.xlsxand place them in thedatafolder - run the Make_FC_data.py
- run the main.py