[1]徐如婷,朱 旗,孙 凯,等.可解释孪生图对比学习及其在脑疾病诊断中的应用[J].南京师大学报(自然科学版),2024,(03):129-137.[doi:10.3969/j.issn.1001-4616.2024.03.016]
　Xu Ruting,Zhu Qi,Sun Kai,et al.Interpretable Siamese Graph Contrastive Networks for Brain Disease Diagnosis[J].Journal of Nanjing Normal University(Natural Science Edition),2024,(03):129-137.[doi:10.3969/j.issn.1001-4616.2024.03.016]
点击复制

可解释孪生图对比学习及其在脑疾病诊断中的应用()
分享到：

《南京师大学报（自然科学版）》[ISSN:1001-4616/CN:32-1239/N]

卷:
期数:
2024年03期
页码:
129-137
栏目:
计算机科学与技术
出版日期:
2024-09-15

文章信息/Info

Title:
Interpretable Siamese Graph Contrastive Networks for Brain Disease Diagnosis
文章编号:
1001-4616(2024)03-0129-09
作者:
徐如婷1; 朱 旗1; 孙 凯2; 朱 敏3; 邵 伟1; 张道强1
(1.南京航空航天大学计算机科学与技术学院,江苏 南京,211106)
(2.深圳市华赛瑞飞智能科技有限公司,广东 深圳 518063)
(3.南京航空航天大学公共实验教学部,江苏 南京 211106)
Author(s):
Xu Ruting1; Zhu Qi1; Sun Kai2; Zhu Min3; Shao Wei1; Zhang Daoqiang1
(1.School of Computer Science and Technology,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China)
(2.Shenzhen Huasai Ruifei Intelligent Technology Co.,Ltd,Shenzhen 518063,China)
(3.Public Experiment Teaching Department of Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China)
关键词:
脑网络研究; 可解释; 图生成网络; 图对比学习; 孪生网络
Keywords:
brain network research; interpretable; graph generation network; graph contrastive learning; siamese networks
分类号:
TP399
DOI:
10.3969/j.issn.1001-4616.2024.03.016
文献标志码:
A
摘要:
近年来,许多研究采用静息态磁共振成像(rs-fMRI)构建动态功能连接网络,并将其应用于癫痫、精神分裂症等脑部疾病的诊断. 研究表明,图卷积网络可以有效保存图结构,提取脑网络的高阶特征. 然而,现有的利用GCN模型对脑疾病进行诊断的模型通常在训练的过程中不进行图结构的更新,因此模型效果常常受到输入图质量的约束. 另外由于GCN模型缺乏可解释性的特点,也限制了模型在疾病诊断上的广泛应用. 在本文中,我们提出了一个可解释的孪生图对比网络用于提取大脑的时空信息. 具体而言,可解释模块通过图生成网络结合脑区之间的相关性信息,动态调整输入模型图结构. 图对比模块学习脑网络全局和局部的信息提高特征表达的鉴别性. 孪生网络模块将成对的样本图作为模型输入,从而缓解数据样本量少的问题. 所有模块联合优化,在个体层面提取大脑网络时域和空域上的特征并进行可靠的疾病诊断. 实验结果表明,该方法在癫痫数据集上具有较好的诊断性能,并为疾病诊断提供可解释性.
Abstract:
In recent years,many studies have used rs-fMRI to construct dynamic functional connectivity network and applied it to the diagnosis of brain diseases such as epilepsy and schizophrenia. The results show that graph convolution network(GCN)can effectively preserve graph structure and extract higher-order features of brain network. However,the existing models using GCN model to diagnose brain diseases usually do not update the graph structure during the training process,so the model effect is often constrained by the quality of the input graph. In addition,the lack of explainability limits the wide application of GCN model in disease diagnosis. In this paper,we propose an interpretable Siamese graph contrastive network to extract spatio-temporal information from the brain network. Specifically,the interpretable module dynamically adjusts the graph structure of the input model by combining the correlation information between brain regions through graph generation network. The graph contrast module learns global and local information of brain network to obtain discriminative feature embedding. Siamese network module takes paired subject graph as model input to alleviate the problem of small data subject size. All modules in our method are optimized together to extract temporal and spatial features of brain networks at the individual level for reliable disease diagnosis. Experimental results show that the proposed method achieves promising diagnostic performance on epilepsy datasets and provides explainable results.

参考文献/References:

[1]FINC K,BONNA K,HE X,et al. Dynamic reconfiguration of functional brain networks during working memory training[J]Nature communications,2020,11(1):1-15.
[2]JIE B,LIU M,SHEN D. Integration of temporal and spatial properties of dynamic connectivity networks for automatic diagnosis of brain disease[J]. Medical image analysis,2018,47:81-94.
[3]WANG M,LIAN C,YAO D,et al. Spatial-temporal dependency modeling and network hub detection for functional MRI analysis via convolutional-recurrent network[J]. IEEE transaction on biomedical engineering,2019,67(8):2241-2252.
[4]YU S,WANG S,XIAO X,et al. Multi-scale enhanced graph convolutional network for early mild cognitive impairment detection[C]//International Conference on Medical Image Computing and Computer-Assisted Intervention. Vancouver,Canada:Springer,Cham,2020:228-237.
[5]ZHANG X,HE L,CHEN K,et al. Multi-view graph convolutional network and its applications on neuroimage analysis for parkinson's disease[J]. AMIA Annual Symposium Proceedings,2018,2018:1147-1156.
[6]ZHU Y,MA J,YUAN C,et al. Interpretable learning based dynamic graph convolutional networks for alzheimer's disease analysis[J]. Information fusion,2022,77:53-61.
[7]CETIN M S,HOUCK J M,RASHID B,et al. Multimodal classification of schizophrenia patients with meg and fmri data using static and dynamic connectivity measures[J]. Frontiers in neuroscience,2016,10:466,.
[8]KOCH G,ZEMEL R,SALAKHUTDINOV R. Siamese neural networks for one-shot image recognition[C]//ICML Deep Learning Workshop. Lille,France:JMLR,2015,2(1):1-30.
[9]SHU H,WANG X,ZHU H T,D-CCA:A decomposition-based canonical correlation analysis for high-dimensional datasets[J]. Journal of the American Statistical Association,2020,115(529):292-306.
[10]YU R,ZHANG H,AN L,et al. Connectivity strength-weighted sparse group representation-based brain network construction form ci classification[J]. Human brain mapping,2017,38(5):2370-2383.
[11]ZHANG M,ZHU Y,WU S,et al. Deep contrastive learning for multi-view network embedding[J/OL]. arXiv preprint arXiv:2108.08296,2021.
[12]HASSANI K,KHASAHMADI A H. Contrastive multi-view representation learning on graphs[C]//International Conference on Machine Learning(ICML). San Diego,CA:PMLR,2020,119(119):4116-4126.
[13]YAN W,ZHANG H,SUI J,et al. Deep chronnectome learning via full bidirectional long short-term memory networks for MCI diagnosis[C]//International Conference on Medical Image Computing and Computer-Assisted Intervention. Granada,Spain:Springer,Cham,2018,11072:249-257.
[14]WANG Q,ZHAN L,THOMPSON P M,et al. Discriminative fusion of multiple brain networks for early mild cognitive impairment detection[C]//2016 IEEE 13th International Symposium on Biomedical Imaging(ISBI). Prague,Czech Republic:IEEE,2016:568-572.
[15]KAWAHARA J,BROWN C J,MILLER S P,et al. BrainNetCNN:convolutional neural networks for brain networks; towards predicting neurodevelopment[J]. NeuroImage,2017,146:1038-1049.
[16]LIU J,MA G,JIANG F,et al. Community-preserving graph convolutions for structural and functional joint embedding of brain networks[C]//2019 IEEE International Conference on Big Data(Big Data). Lòs angeles,CA:IEEE,2019:1163-1168.
[17]MCDANIEL C,QUINN S. Developing a graph convolution-based analysis pipeline for multi-modal neuroimage data:an application to parkinson's disease[C]//Proceedings of the 18th Python in Science Conference(SciPy 2019). Austin TX:IEEE 2019:42-49.
[18]HUANG J,ZHOU L,WANG L,et al. Attention-diffusion-bilinear neural network for brain network analysis[J]. IEEE transactions on medical imaging,2020,39(7):2541-2552.
[19]WAITES A B,BRIELLMANN R S,SALING M M,et al. Functional connectivity networks are disrupted in left temporal lobe epilepsy[J]. Annals of neurology,2006,59(2):335-343.
[20]Braakman H M,Vaessen M J,Jansen J F,et al. Frontal lobe connectivity and cognitive impairment in pediatric frontal lobe epilepsy[J]. Epilepsia,2013,54(3):446-454.
[21]STRUB R L. Frontal lobe syndrome in a patient with bilateral globus pallidus lesions[J]. Archives of neurology,1989,46(9):1024-1027.
[22]GLOOR P,OLIVIER A,QUESNEY L F,et al. The role of the limbic system in experiential phenomena of temporal lobe epilepsy[J]. Annals of neurology,1982,12(2):129-144.

备注/Memo

备注/Memo:
收稿日期:2022-09-20.
基金项目:江苏省自然科学基金项目(BK20231438)、江苏省重点研发项目(BE202282)、国家自然科学基金项目(62076129、62371234).
通讯作者:朱旗,博士,副教授,研究方向:机器学习,模式识别和脑疾病诊断. E-mail:zhuqinuaa@163.com
更新日期/Last Update: 2024-09-15