[1]陈娜,张冰涛.面向脑功能网络的张量分解及抑郁症分析[J].中国医学物理学杂志,2026,43(9):1225-1235.[doi:DOI:10.3969/j.issn.1005-202X.2026.09.010]
　CHEN Na,ZHANG Bingtao.Tensor decomposition for brain functional networks and depression analysis[J].Chinese Journal of Medical Physics,2026,43(9):1225-1235.[doi:DOI:10.3969/j.issn.1005-202X.2026.09.010]
点击复制

面向脑功能网络的张量分解及抑郁症分析()

《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]

卷:
43卷
期数:
2026年第9期
页码:
1225-1235
栏目:
医学信号处理与医学仪器
出版日期:
2026-09-30

文章信息/Info

Title:
Tensor decomposition for brain functional networks and depression analysis
文章编号:
1005-202X（2026）09-1225-11
作者:
陈娜; 张冰涛
兰州交通大学电子与信息工程学院， 甘肃 兰州 730070
Author(s):
CHEN Na;  ZHANG Bingtao
School of Electronic and Information Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China
关键词:
脑功能网络; 经验模态分解; 快速耦合非负张量分解; 抑郁症
Keywords:
Keywords: brain functional network empirical mode decomposition fast coupled non-negative tensor decomposition depression
分类号:
R318；TP391
DOI:
DOI:10.3969/j.issn.1005-202X.2026.09.010
文献标志码:
A
摘要:
基于脑电的抑郁症（DP）分析往往将电极视为孤立节点，忽略节点之间的关联性。为此，提出一种时-频-空信息融合的快速耦合非负张量分解（F-CNTD）模型，以揭示DP患者脑功能网络（BFN）的异常连接变化。引入恒定信噪比与新型局部均值算子改进集合经验模态分解（EEMD），有效抑制模态混叠，提取稳定的固有模态函数（IMF）。基于相位滞后指数构建功能连接张量，并结合耦合约束与快速分层最小二乘法实现F-CNTD分解。实验结果表明，改进EEMD在抑制模态混叠和提高分解稳定性方面优于EEMD和CEEMDAN；F-CNTD分析发现，DP患者在多个IMF分量上的全脑连接强度显著弱于正常对照组，左额颞叶区域聚集系数与局部效率显著下降，小世界属性在全脑区存在显著组间差异，且Hub节点在IMF3和IMF6中呈现左右分布不对称性。此外，本研究方法能有效识别DP相关的BFN异常，为DP客观诊断提供新途径。
Abstract:
Abstract: Given that the electroencephalogram-based analysis of depression often treats electrodes as isolated nodes, neglecting the interconnections between nodes, a fast coupled non-negative tensor decomposition (F-CNTD) model that integrates time-frequency-spatial information is developed to uncover abnormal connectivity changes in brain functional networks of depression patients. Specifically, a constant signal-to-noise ratio and a novel local mean operator are introduced to improve the ensemble empirical mode decomposition (EEMD), effectively suppressing mode mixing and extracting stable intrinsic mode functions (IMF). A functional connectivity tensor is constructed based on the phase lag index, and the F-CNTD decomposition is implemented using coupling constraints and the fast hierarchical alternating least squares algorithm. Experimental results show that the improved EEMD outperforms EEMD and CEEMDAN in terms of mode mixing suppression and decomposition stability. F-CNTD analysis demonstrates that depression patients exhibit significantly weaker whole-brain connectivity strength across multiple IMF components compared with normal controls, with a significant decrease in clustering coefficient and local efficiency in the left frontal-temporal region. Significant inter-group differences in small-world property across the whole brain are also observed. Additionally, the distribution of Hub nodes show left-right asymmetry in IMF3 and IMF6. The findings show that the proposed method can effectively identify brain functional network abnormalities related to depression, providing a new avenue for objective diagnosis of depression.

相似文献/References:

[1]韩庆阳,李丙玉,王晓东,等.一种同时消除脉搏波信号中呼吸基线漂移和高频噪声的方法[J].中国医学物理学杂志,2014,31(02):4801.[doi:10.3969/j.issn.1005-202X.2014.02.019]
[2]田树香,朱健铭,陈真诚,等.一种新的人体呼吸波采集方法与实现[J].中国医学物理学杂志,2014,31(05):5169.[doi:10.3969/j.issn.1005-202X.2014.05.016]
[3]郑威,于洪丽,丁卫国,等. 基于格兰杰因果关系磁刺激穴位对亚健康失眠的脑功能网络分析[J].中国医学物理学杂志,2018,35(2):228.[doi:DOI:10.3969/j.issn.1005-202X.2018.02.022]
　ZHENG Wei,YU Hongli,DING Weiguo,et al. Brain functional network analysis of magnetic stimulation at acupoints for sub-health insomnia based on Granger causality[J].Chinese Journal of Medical Physics,2018,35(9):228.[doi:DOI:10.3969/j.issn.1005-202X.2018.02.022]
[4]冯凯,董秀成,刘栋博.基于经验模态分解-小波包变换的表面肌电信号手势识别[J].中国医学物理学杂志,2021,38(4):461.[doi:DOI:10.3969/j.issn.1005-202X.2021.04.013]
　FENG Kai,DONG Xiucheng,LIU Dongbo.Empirical mode decomposition and wavelet packet transform applied to surface EMG signal for hand gesture recognition[J].Chinese Journal of Medical Physics,2021,38(9):461.[doi:DOI:10.3969/j.issn.1005-202X.2021.04.013]
[5]黄益平,陈真诚,梁永波,等.基于脉搏波的呼吸频率计算方法[J].中国医学物理学杂志,2021,38(9):1124.[doi:10.3969/j.issn.1005-202X.2021.09.014]
　HUANG Yiping,CHEN Zhencheng,LIANG Yongbo,et al.Calculation methods of respiratory rate based on PPG[J].Chinese Journal of Medical Physics,2021,38(9):1124.[doi:10.3969/j.issn.1005-202X.2021.09.014]
[6]孙文慧,张乙鹏,林冬梅,等.基于变分模态分解的肺音去噪算法[J].中国医学物理学杂志,2024,41(4):479.[doi:DOI:10.3969/j.issn.1005-202X.2024.04.013]
　SUN Wenhui,ZHANG Yipeng,et al.Lung sound denoising algorithm based on variational mode decomposition[J].Chinese Journal of Medical Physics,2024,41(9):479.[doi:DOI:10.3969/j.issn.1005-202X.2024.04.013]
[7]王建尚,张冰涛,王小敏,等.基于频空融合与3D-CNN-Attention的抑郁症识别[J].中国医学物理学杂志,2024,41(10):1307.[doi:DOI:10.3969/j.issn.1005-202X.2024.10.016]
　WANG Jianshang,ZHANG Bingtao,WANG Xiaomin,et al.Depression recognition based on frequency-space domain fusion and 3D-CNN-Attention[J].Chinese Journal of Medical Physics,2024,41(9):1307.[doi:DOI:10.3969/j.issn.1005-202X.2024.10.016]
[8]张培玲,候康.基于多尺度自适应残差网络的癫痫检测方法[J].中国医学物理学杂志,2025,42(3):381.[doi:10.3969/j.issn.1005-202X.2025.03.015]
　ZHANG Peiling,HOU Kang.Epilepsy detection method based on multi-scale adaptive residual network[J].Chinese Journal of Medical Physics,2025,42(9):381.[doi:10.3969/j.issn.1005-202X.2025.03.015]
[9]薛栋超,朱俊杰.针对心磁信号的HISSA优化多模态协同降噪算法[J].中国医学物理学杂志,2025,42(9):1201.[doi:DOI:10.3969/j.issn.1005-202X.2025.09.012]
　XUE Dongchao,ZHU Junjie,et al.HISSA-optimized multi-level cooperative denoising algorithm for magnetocardiogram signals[J].Chinese Journal of Medical Physics,2025,42(9):1201.[doi:DOI:10.3969/j.issn.1005-202X.2025.09.012]

备注/Memo

备注/Memo:
【收稿日期】2026-03-05 【基金项目】国家自然科学基金（62362047，61962034） 【作者简介】陈娜，硕士研究生，研究方向：脑电信号分析、多模态融合，E-mail: 13439372149@163.com 【通信作者】张冰涛，博士生导师，教授，研究方向：多模态融合，E-mail: zhangbingtao321@163.com
更新日期/Last Update: 2026-09-30