The rapid accumulation of electronic health records (EHRs) and the advancements in data analysis technology have laid the foundation for research and clinical decision-making in the healthcare community. Graph neural networks (GNNs), a deep learning model family for graph embedding representations, have been widely used in the field of smart healthcare. However, traditional GNNs rely on the basic assumption that the graph structure extracted from the complex interactions among the EHRs must be a real topology. Noisy connections or false topology in the graph structure leads to inefficient disease prediction. We devise a new model named PM-GSL to improve diabetes clinical assistant diagnosis based on patient multi-relational graph structure learning. Specifically, we first build a patient multi-relational graph based on patient demographics, diagnostic information, laboratory tests, and complex interactions between medicines in EHRs. Second, to fully consider the heterogeneity of the patient multi-relational graph, we consider the node characteristics and the higher-order semantics of nodes. Thus, three candidate graphs are generated in the PM-GSL model: original subgraph, overall feature graph, and higher-order semantic graph. Finally, we fuse the three candidate graphs into a new heterogeneous graph and jointly optimize the graph structure with GNNs in the disease prediction task. The experimental results indicate that PM-GSL outperforms other state-of-the-art models in diabetes clinical assistant diagnosis tasks.
Citation: Yong Li, Li Feng. Patient multi-relational graph structure learning for diabetes clinical assistant diagnosis[J]. Mathematical Biosciences and Engineering, 2023, 20(5): 8428-8445. doi: 10.3934/mbe.2023369
The rapid accumulation of electronic health records (EHRs) and the advancements in data analysis technology have laid the foundation for research and clinical decision-making in the healthcare community. Graph neural networks (GNNs), a deep learning model family for graph embedding representations, have been widely used in the field of smart healthcare. However, traditional GNNs rely on the basic assumption that the graph structure extracted from the complex interactions among the EHRs must be a real topology. Noisy connections or false topology in the graph structure leads to inefficient disease prediction. We devise a new model named PM-GSL to improve diabetes clinical assistant diagnosis based on patient multi-relational graph structure learning. Specifically, we first build a patient multi-relational graph based on patient demographics, diagnostic information, laboratory tests, and complex interactions between medicines in EHRs. Second, to fully consider the heterogeneity of the patient multi-relational graph, we consider the node characteristics and the higher-order semantics of nodes. Thus, three candidate graphs are generated in the PM-GSL model: original subgraph, overall feature graph, and higher-order semantic graph. Finally, we fuse the three candidate graphs into a new heterogeneous graph and jointly optimize the graph structure with GNNs in the disease prediction task. The experimental results indicate that PM-GSL outperforms other state-of-the-art models in diabetes clinical assistant diagnosis tasks.
| [1] | M. Lin, S. Wang, Y. Ding, An empirical study of using radiology reports and images to improve ICU-mortality prediction, in 2021 IEEE 9th International Conference on Healthcare Informatics (ICHI), (2021), 497–498. https://doi.org/10.1109/ICHI52183.2021.00088 |
| [2] |
Z. Liang, Z. Zhang, H. Chen, Disease prediction based on multi-type data fusion from Chinese electronic health record, Math. Biosci. Eng., 19 (2022), 13732–13746. https://doi.org/10.3934/mbe.2022640 doi: 10.3934/mbe.2022640
|
| [3] | Z. Wang, R. Wen, X. Chen, Online disease diagnosis with inductive heterogeneous graph convolutional networks, in Proceedings of the Web Conference 2021, (2021), 3349–3358. https://doi.org/10.1145/3442381.3449795 |
| [4] | Z. Liu, X. Li, H. Peng, Heterogeneous similarity graph neural network on electronic health records, in 2020 IEEE International Conference on Big Data (Big Data), (2020), 1196–1205. https://doi.org/10.1109/BigData50022.2020.9377795 |
| [5] |
F. Scarselli, M. Gori, A. C. Tsoi, The graph neural network model, IEEE Trans. Neural Networks, 20 (2008), 61–80. https://doi.org/10.1109/TNN.2008.2005605 doi: 10.1109/TNN.2008.2005605
|
| [6] | M. Defferrard, X. Bresson, P. Vandergheynst, Convolutional neural networks on graphs with fast localized spectral filtering, in Advances in Neural Information Processing Systems, 29 (2016). |
| [7] | T. N. Kipf, M. Welling, Semi-supervised classification with graph convolutional networks, preprint, arXiv: 1609.02907. |
| [8] | F. Wu, A. Souza, T. Zhang, Simplifying graph convolutional networks, in International Conference on Machine Learning, (2019), 6861–6871. |
| [9] | J. Zhang, X. Shi, J. Xie, Gaan: Gated attention networks for learning on large and spatiotemporal graphs, preprint, arXiv: 1803.07294. |
| [10] |
K. Zhang, B. Hu, F. Zhou, Graph-based structural knowledge-aware network for diagnosis assistant, Math. Biosci. Eng., 19 (2022), 10533–10549. https://doi.org/10.3934/mbe.2022492 doi: 10.3934/mbe.2022492
|
| [11] | W. L. Hamilton, R. Ying, J. Leskovec, Inductive representation learning on large graphs, in Proceedings of the 31st International Conference on Neural Information Processing Systems, (2017), 1025–1035. |
| [12] | P. Veličković, G. Cucurull, A. Casanova, Graph attention networks, preprint, arXiv: 1710.10903. |
| [13] | Y. Dong, N. V. Chawla, A. Swami, metapath2vec: Scalable representation learning for heterogeneous networks, in Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, (2017), 135–144. https://doi.org/10.1145/3097983.3098036 |
| [14] | X. Wang, H. Ji, C. Shi, Heterogeneous graph attention network, in The World Wide Web Conference, (2019), 2022–2032. https://doi.org/10.1145/3308558.3313562 |
| [15] | X. Fu, J. Zhang, Z. Meng, Magnn: Metapath aggregated graph neural network for heterogeneous graph embedding, in Proceedings of The Web Conference 2020, (2020), 2331–2341. https://doi.org/10.1145/3366423.3380297 |
| [16] | X. Wang, N. Liu, H. Han, Self-supervised heterogeneous graph neural network with co-contrastive learning, in Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, (2021), 1726–1736. https://doi.org/10.1145/3447548.3467415 |
| [17] |
M. E. J. Newman, Estimating network structure from unreliable measurements, Phys. Rev. E, 98 (2018), 62321. https://doi.org/10.1103/PhysRevE.98.062321 doi: 10.1103/PhysRevE.98.062321
|
| [18] |
T. Martin, B. Ball, M. E. J. Newman, Structural inference for uncertain networks, Phys. Rev. E, 93 (2016), 12306. https://doi.org/10.1103/PhysRevE.93.012306 doi: 10.1103/PhysRevE.93.012306
|
| [19] |
W. X. Wan, Y. C. Lai, C. Grebogi, Data based identification and prediction of nonlinear and complex dynamical systems, Phys. Rep., 644 (2016), 1–76. https://doi.org/10.1016/j.physrep.2016.06.004 doi: 10.1016/j.physrep.2016.06.004
|
| [20] | L. Franceschi, M. Niepert, M. Pontil, Learning discrete structures for graph neural networks, in International Conference on Machine Learning, (2019), 1972–1982. |
| [21] | W. Jin, Y. Ma, X. Liu, Graph structure learning for robust graph neural networks, in Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, (2020), 66–74. https://doi.org/10.1145/3394486.3403049 |
| [22] | S. Yun, M. Jeong, R. Kim, Graph transformer networks, in Advances in Neural Information Processing Systems, 32 (2019). |
| [23] | Y. Ji, G. Chu, X. Wang, Prohibited item detection via risk graph structure learning, in Proceedings of the ACM Web Conference 2022, (2022), 1434–1443. https://doi.org/10.1145/3485447.3512190 |
| [24] |
G. Wen, P. Cao, H. Bao, MVS-GCN: A prior brain structure learning-guided multi-view graph convolution network for autism spectrum disorder diagnosis, Comput. Biol. Med., 142 (2022), 105239. https://doi.org/10.1016/j.compbiomed.2022.105239 doi: 10.1016/j.compbiomed.2022.105239
|
| [25] | S. Tang, A. Tariq, J. Dunnmon, Multimodal spatiotemporal graph neural networks for improved prediction of 30-day all-cause hospital readmission, preprint, arXiv: 2204.06766. |
| [26] |
S. Zheng, Z. Zhu, Z. Liu, Multi-modal graph learning for disease prediction, IEEE Trans. Med. Imaging, 41 (2022), 2207–2216. https://doi.org/10.1109/TMI.2022.3159264 doi: 10.1109/TMI.2022.3159264
|
| [27] | E. Choi, Z. Xu, Y. Li, Learning the graphical structure of electronic health records with graph convolutional transformer, in Proceedings of the AAAI Conference on Artificial Intelligence, 34 (2020), 606–613. https://doi.org/10.1609/aaai.v34i01.5400 |
| [28] | Y. Cao, H. Peng, P. S. Yu, Multi-information source hin for medical concept embedding, in Pacific-Asia Conference on Knowledge Discovery and Data Mining, Springer, Cham, (2020), 396–408. https://doi.org/10.1007/978-3-030-47436-2_30 |
| [29] | J. Gilmer, S. S. Schoenholz, P. F. Riley, Neural message passing for quantum chemistry, in International Conference on Machine Learning, (2017), 1263–1272. |
| [30] | J. Zhao, X. Wang, C. Shi, Heterogeneous graph structure learning for graph neural networks, in Proceedings of the AAAI Conference on Artificial Intelligence, 35 (2021), 4697–4705. https://doi.org/10.1609/aaai.v35i5.16600 |
Figures(6) / Tables(4)
Yong Li, Li Feng. Patient multi-relational graph structure learning for diabetes clinical assistant diagnosis[J]. Mathematical Biosciences and Engineering, 2023, 20(5): 8428-8445. doi: 10.3934/mbe.2023369
DownLoad: