This is the official PyTorch implementation of BrainPrompt+ from the paper " BrainPrompt+: Multi-Level Brain Prompt Learning for Knowledge-Guided Neurological Disorder Identification" accepted by IEEE Transactions on Medical Imaging (TMI'26).
Paper Link: IEEE Xplore
All Preprocessed data used in this paper are published in this paper.
Data splits and configurations are stored in ./data/ and ./configs/. If you want to process your own data, please check the dataloader script ./data/BrainNet.py.
Please check baseline.sh on how to run the project.
If you find this code useful, please consider citing our paper:
@ARTICLE{11518539,
author={Xu, Jiaxing and He, Kai and Tang, Yue and Li, Wei and Lan, Mengcheng and Xun, Yue and Lin, Qika and Ran, Peifan and Ke, Yiping and Feng, Mengling},
journal={IEEE Transactions on Medical Imaging}, title={BrainPrompt+: Multi-Level Brain Prompt Learning for Knowledge-Guided Neurological Disorder Identification}, year={2026},
volume={},
number={},
pages={1-1},
keywords={Brain;Modeling;Diseases;Biomedical imaging;Graph neural networks;Learning (artificial intelligence);Functional magnetic resonance imaging;Accuracy;Conferences;Labeling;Brain Network;fMRI;Graph Neural Network;Neurological Disorder;Prompt Enhancement},
doi={10.1109/TMI.2026.3692958}}
If you have any questions, please feel free to reach out at jiaxing003@e.ntu.edu.sg.