FuzzyCAL: A Fuzzy-Logic Enhanced Causal Attention GNN for Robust Cocaine Use Disorder Classification

Document Type : Research Paper

Author

Electrical and Computer engineering department, Tarbiat Modares University

Abstract

Fuzzy logic has emerged as a powerful tool for handling uncertainty and imprecision in machine learning, enabling models to make robust predictions in complex, noisy domains. In this work, we integrate fuzzy-set theory with causal attention learning to improve the classification of Cocaine Use Disorder (CUD) from resting-state functiona Magnetic Resonance Imaging (fMRI) brain networks. Building on the Causal Attention Learning (CAL) framework, we introduce FuzzyCAL, which augments node- and edge-level causal masks with fuzzy membership functions that quantify “weak,” “medium,” and “strong” relevance of graph components. This fusion of fuzzy logic and causal Graph Neural Networks (GNNs) not only preserves the interpretability of causal attention but also adapts to the inherent variability of neuroimaging data. Through extensive 5-fold cross-validation on the SUDMEX
CONN dataset, FuzzyCAL achieves a mean test accuracy of 86.20% (±7.18%) and F1-score of 85.87% (±7.23%), outperforming both a non-causal GNN baseline and the original CAL model. We further demonstrate how fuzzy-weighted causal masks reveal anatomically meaningful biomarkers in temporal and sensorimotor cortices. Our results suggest that embedding fuzzy reasoning into causal graph models enhances both predictive performance and neuroscientific interpretability, offering a promising direction for precision diagnostics in substance use disorders.

Keywords

Main Subjects

References

[1] M. Alavi, A. Valiollahi, M. Pakravan, Bibliometric analysis of research trends on graph neural networks, 0th CSI
International Symposium on Artificial Intelligence and Signal Processing (AISP), (2024), 1-8. https://doi.org/
10.1109/AISP61396.2024.10475285
[2] D. Angeles-Valdez, J. Rasgado-Toledo, V. Issa-Garcia, T. Balducci, V. Villica˜na, A. Valencia, J. J. Gonzalez-
Olvera, E. Reyes-Zamorano, E. A. Garza-Villarreal, The Mexican magnetic resonance imaging dataset of patients
with cocaine use disorder: SUDMEX CONN, Scientific Data, 9(1) (2022), 133. https://doi.org/10.1038/
s41597-022-01251-3
[3] S. Bastami, R. Pirmohamadiani, M. Dowlatshahi, A. Abdollahpouri, Enhanced high-dimensional data classification
by combining fuzzy learning integration and graph transformers, Iranian Journal of Fuzzy Systems, 22(2) (2025),
1-15. https://doi.org/10.22111/ijfs.2025.50945.8999
[4] E. Bullmore, O. Sporns, Complex brain networks: Graph theoretical analysis of structural and functional systems,
Nature Reviews Neuroscience, 10(3) (2009), 186-198. https://doi.org/10.1038/nrn2575
[5] J. M. Cisler, A. Elton, A. P. Kennedy, J. Young, S. Smitherman, G. A. James, C. D. Kilts, Altered functional connectivity
of the insular cortex across prefrontal networks in cocaine addiction, Psychiatry Research: Neuroimaging,
213(1) (2013), 39-46. https://doi.org/10.1016/j.pscychresns.2013.02.007
[6] R. W. Cox, AFNI: Software for analysis and visualization of functional magnetic resonance neuroimages, Computers
and Biomedical Research, 29(3) (1996), 162-173. https://doi.org/10.1006/cbmr.1996.0014
[7] Y. Gao, L. Shen, S. T. Xia, Dag-gan: Causal structure learning with generative adversarial nets, 2021 IEEE
International Conference on Acoustics, Speech and Signal Processing, (2021), 3320-3324. https://doi.org/10.
1109/ICASSP39728.2021.9414770
[8] M. Gholami, M. Faramarzi, N. Alipour, M. Pakravan, Node embedding extraction for causal brain graphs in fMRI
data, Proc. 2025 29th International Computer Conference, Computer Society of Iran (CSICC), (2025), 1-6. https:
//doi.org/10.1109/CSICC65765.2025.10967434
[9] A. A. Hagberg, D. A. Schult, P. J. Swart, Exploring network structure, dynamics, and function using network X,
Proc. 7th Python in Science Conference (SciPy 2008), (2008), 11-15.https://doi.org/10.25080/tcwv9851
[10] T. R. Khalifa, X. Yu, A. Sharafian, X. Zhong, Z. Wu, Interval type-3 fuzzy Wiener model for nonlinear dynamic
systems: Application to continuous stirred tank reactor, Chaos, Solitons and Fractals, 199 (2025), 116584. https:
//doi.org/10.1016/j.ins.2021.09.004
[11] T. R. Khalifa, X. Yu, X. Zhong, Z. Wu, Adaptive general type-2 fuzzy model-based control for nonlinear networked
systems with packet dropouts, ISA Transactions, 159 (2025), 257-277. https://doi.org/10.1016/j.isatra.
2025.02.009
[12] T. R. Khalifa, X. Yu, X. Zhong, Z. Wu, Indirect adaptive interval type-3 fuzzy tracking control for nonlinear
discrete-time networked control systems with DoS attacks, IEEE Transactions on Cybernetics, (2025). https:
//doi.org/10.1109/TCYB.2025.3591555
[13] P. Li, X. Wang, Z. Zhang, Z. Zhang, F. Shen, J. Wang, Y. Li, W. Zhu, Causal-aware graph neural architecture
search under distribution shifts, Proc. 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining,
2 (2025), 1458-1469. https://doi.org/10.1145/3711896.3736873
[14] Y. N. Lin, H. C. Cai, C. Y. Zhang, H. Y. Yao, C. L. P. Chen, Fuzzy neural network for representation learning
on uncertain graphs, IEEE Transactions on Fuzzy Systems, 32(9) (2024), 5259-5271. https://doi.org/10.1109/
TFUZZ.2024.3418902
[15] Y. Lin, H. Dong, H. Wang, T. Zhang, Bayesian invariant risk minimization, Proc. IEEE/CVF Conference on
Computer Vision and Pattern Recognition (CVPR), (2022), 16021-16030. https://doi.org/10.1109/CVPR52688.
2022.01555
[16] C. Lu, Y. Wu, J. M. Hern´andez-Lobato, B. Sch¨olkopf, Invariant causal representation learning for out-ofdistribution
generalization, In International Conference on Learning Representations, (2021).
[17] X. Ma, L. Chen, Z. Deng, P. Xu, Q. Yan, K. S. Choi, S. Wang, Deep image feature learning with fuzzy rules,
IEEE Transactions on Emerging Topics in Computational Intelligence, 8(1) (2023), 724-737. https://doi.org/
10.1109/TETCI.2023.3259447
[18] M. Pakravan, M. Abbaszadeh, A. Ghazizadeh, Coordinated multivoxel coding beyond univariate effects is not likely
to be observable in fMRI data, NeuroImage, 247 (2022), 118825. https://doi.org/10.1016/j.neuroimage.2021.
118825
[19] M. Pakravan, M. B. Shamsollahi, Joint, partially-joint, and individual independent component analysis in multisubject
fMRI data, IEEE Transactions on Biomedical Engineering, 67(7) (2020), 1969-1981. https://doi.org/
10.1109/TBME.2019.2953274
[20] J. Peters, D. Janzing, B. Sch¨olkopf, Elements of causal inference: Foundations and learning algorithms, MIT Press,
(2017). https://doi.org/10.5555/3202377
[21] A. A. Pollard, A. O. Hauson, N. S. Lackey, E. Zhang, S. Khayat, B. Carson, L. Fortea, J. Radua, I. Grant,
Functional neuroanatomy of craving in heroin use disorder: Voxel-based meta-analysis of functional magnetic
resonance imaging (fMRI) drug cue reactivity studies, The American Journal of Drug and Alcohol Abuse, 49(4)
(2023), 418-430. https://doi.org/10.1080/00952990.2023.2172423
[22] J. Qiu, J. Xie, D. Zhang, R. Zhang, M. Lin, A fuzzy twin support vector machine based on dissimilarity measure
and its biomedical applications, International Journal of Fuzzy Systems, 26(8) (2024), 2750-2766. https://doi.
org/10.1007/s40815-024-01725-z
[23] T. J. Ross, Fuzzy logic with engineering applications, 3rd ed., John Wiley and Sons, (2010). https://doi.org/
10.1002/9781119994374
[24] B. Sch¨olkopf, J. Peters, D. Janzing, J. Zscheischler, K. Zhang, Towards causal representation learning, Proceedings
of the IEEE, 109(5) (2021), 612-634. https://doi.org/10.48550/arXiv.2102.11107
[25] R. Shalaby, M. Ibrahim, T. Khalifa, Fuzzy-supervised PI for flow-control orocess, design and experimental validation,
Proc. 2019 Novel Intelligent and Leading Emerging Sciences Conference (NILES), (2019), 249-253.
https://doi.org/10.1109/NILES.2019.8909327
[26] Y. Sui, X. Wang, J. Wu, M. Lin, X. He, T. S. Chua, Causal attention for interpretable and generalizable graph
classification, Proc. 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, (2022), 1696-1705.
https://doi.org/10.1145/3534678.3539366
[27] L. Wang, M. Zhang, et al., A brain structure learning-guided multi-view graph representation learning for brain
network analysis, IEEE Transactions on Medical Imaging, 42(5) (2023), 1235-1247. https://doi.org/10.21037/
qims-24-578
[28] K. Xu, W. Hu, J. Leskovec, S. Jegelka, How powerful are graph neural networks?, International Conference on
Learning Representations (ICLR), (2019). https://doi.org/10.48550/arXiv.1810.00826
[29] H. Yu, X. Lei, Z. Song, C. Liu, J. Wang, Supervised network-based fuzzy learning of EEG signals for Alzheimer’s
disease identification, IEEE Transactions on Fuzzy Systems, 28(1) (2019), 60-71. https://doi.org/10.1109/
TFUZZ.2019.290375
[30] L. A. Zadeh, Fuzzy sets, Information and Control, 8(3) (1965), 338-353. https://doi.org/10.1016/
S0019-9958(65)90241-X
[31] K. Zhao, G. A. Fonzo, H. Xie, D. J. Oathes, C. J. Keller, N. Carlisle, A. Etkin, E. A. Garza-Villarreal, Y.
Zhang, A generalizable functional connectivity signature characterizes brain dysfunction and links to rTMS treatment
response in cocaine use disorder, Nature Mental Health, 2(4) (2024), 388-400. https://doi.org/10.1038/
s44220-024-00209-1
Iranian Journal of Fuzzy Systems
Volume 22, Issue 6
November and December 2025
Pages 167-182

Files

Share

How to cite

Statistics