[1]于洪仕,林悦,李宏宇.基于混合注意力Transformer的癫痫检测方法[J].中国医学物理学杂志,2026,43(6):811-817.[doi:DOI:10.3969/j.issn.1005-202X.2026.06.015]
　YU Hongshi,LIN Yue,et al.Epilepsy detection method based on hybrid attention Transformer[J].Chinese Journal of Medical Physics,2026,43(6):811-817.[doi:DOI:10.3969/j.issn.1005-202X.2026.06.015]
点击复制

基于混合注意力Transformer的癫痫检测方法()

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

卷:
43卷
期数:
2026年第6期
页码:
811-817
栏目:
医学影像物理
出版日期:
2026-06-26

文章信息/Info

Title:
Epilepsy detection method based on hybrid attention Transformer
文章编号:
1005-202X（2026）06-0811-07
作者:
于洪仕1; 2; 林悦1; 李宏宇1
1.辽宁工程技术大学电子与信息工程学院， 辽宁 葫芦岛 125105； 2.辽宁工程技术大学辽宁省射频大数据智能应用重点实验室，辽宁 葫芦岛 125105
Author(s):
YU Hongshi 1;  2;  LIN Yue1;  LI Hongyu1
1. School of Electronic and Information Engineering, Liaoning Technical University, Huludao 125105, China 2. Liaoning Key Laboratory of Radio Frequency Big Data Intelligent Application, Liaoning Technical University, Huludao 125105, China
关键词:
脑电信号; 癫痫; 张量; 混合注意力; Transformer
Keywords:
Keywords: electroencephalogram signal epilepsy tensor hybrid attention Transformer
分类号:
R318；TP391
DOI:
DOI:10.3969/j.issn.1005-202X.2026.06.015
文献标志码:
A
摘要:
针对现有癫痫检测方法的输入信号为单一特征，无法充分提取多维特征问题，提出一种融合频域、非线性域和时域脑电信号特征的多维特征张量，分别是相位锁相值、互信息和肯德尔相关系数[τb]，并精心设计一个与之相匹配的混合注意力Transformer模型（HA-Trans），该模型巧妙融合空间注意力、交叉注意力和时间注意力机制，充分提取多维脑电特征，最终输入到全连接层实现分类。本文方法在CHB-MIT数据集的分类准确率为99.17%，特异性为99.03%，敏感性为98.95%，优于现有方法，实现癫痫的有效检测。
Abstract:
Abstract: Given that existing epilepsy detection methods adopt single-dimensional features as inputs and fail to fully extract multidimensional features, this study proposes a multidimensional feature tensor which integrates frequency-, nonlinear-, and time-domain electroencephalogram features: phase locking value, mutual information, and Kendall correlation coefficient [τb]. Accordingly, a hybrid attention Transformer (HA-Trans) model is developed to effectively extract multidimensional EEG features via spatial attention, cross attention, and temporal attention mechanisms, and the extracted features are then input into the fully connected layer for classification. The proposed method achieves a classification accuracy of 99.17%, specificity of 99.03%, and sensitivity of 98.95% on the CHB-MIT dataset, outperforming existing methods and enabling effective epilepsy detection.

相似文献/References:

[1]王怡玲,覃玉荣,郭湛超,等.基于不同闪烁频率光刺激的脑电压变化研究[J].中国医学物理学杂志,2014,31(05):5184.[doi:10.3969/j.issn.1005-202X.2014.05.019]
[2]李凯,姜永涛,于海波,等.癫痫治疗效果的复杂度评估分析[J].中国医学物理学杂志,2013,30(03):4169.[doi:10.3969/j.issn.1005-202X.2013.03.019]
[3]杨建平,张德乾,吕敬祥,等.操作发起过程多脑区协作的脑电谱熵特征[J].中国医学物理学杂志,2016,33(1):44.[doi:DOI:10.3969/j.issn.1005-202X.2016.01.010]
　[J].Chinese Journal of Medical Physics,2016,33(6):44.[doi:DOI:10.3969/j.issn.1005-202X.2016.01.010]
[4]顾家军,叶继伦.麻醉深度监测中脑电信号特征提取方法[J].中国医学物理学杂志,2016,33(2):157.[doi:10.3969/j.issn.1005-202X.2016.02.010]
　[J].Chinese Journal of Medical Physics,2016,33(6):157.[doi:10.3969/j.issn.1005-202X.2016.02.010]
[5]刘岩,李幼军,陈萌. 基于固有模态分解和深度学习的抑郁症脑电信号分类分析[J].中国医学物理学杂志,2017,34(9):963.[doi:DOI:10.3969/j.issn.1005-202X.2017.09.021]
　[J].Chinese Journal of Medical Physics,2017,34(6):963.[doi:DOI:10.3969/j.issn.1005-202X.2017.09.021]
[6]李春,黄波,唐海宁,等. 癫痫伴发焦虑抑郁患者脑神经递质活动的脑电超慢涨落图表现分析[J].中国医学物理学杂志,2018,35(3):369.[doi:DOI:10.3969/j.issn.1005-202X.2018.03.022]
　LI Chun,HUANG Bo,TANG Haining,et al. Characteristics of intracerebral neurotransmitter activity on encephalofluctuograph in epileptic patients with anxiety and depression[J].Chinese Journal of Medical Physics,2018,35(6):369.[doi:DOI:10.3969/j.issn.1005-202X.2018.03.022]
[7]马玉良,刘卫星,张淞杰,等.基于ABC-SVM的运动想象脑电信号模式分类[J].中国医学物理学杂志,2018,35(9):1056.[doi:10.3969/j.issn.1005-202X.2018.09.012]
　MAYuliang,LIUWeixing,ZHANG Songjie,et al.Pattern classification of motor imagery EEG signals based on ABC-SVM algorithm[J].Chinese Journal of Medical Physics,2018,35(6):1056.[doi:10.3969/j.issn.1005-202X.2018.09.012]
[8]刘畅,覃玉荣,时文健.视听觉刺激下大脑头皮电位空间变化特性[J].中国医学物理学杂志,2018,35(10):1225.[doi:DOI:10.3969/j.issn.1005-202X.2018.010.023]
　LIU Chang,QIN Yurong,SHI Wenjian. Spatial variation characteristics of scalp potentials under audiovisual stimuli[J].Chinese Journal of Medical Physics,2018,35(6):1225.[doi:DOI:10.3969/j.issn.1005-202X.2018.010.023]
[9]周杰,杨国雨,徐涛. 基于空间频率与时间序列信息的多类运动想象脑电分类[J].中国医学物理学杂志,2019,36(1):81.[doi:DOI:10.3969/j.issn.1005-202X.2019.01.016]
　ZHOU Jie,YANG Guoyu,XU Tao. Classification of multi-class motor imagery EEG data based on spatial frequency and time-series information[J].Chinese Journal of Medical Physics,2019,36(6):81.[doi:DOI:10.3969/j.issn.1005-202X.2019.01.016]
[10]李冬,金韬,冯智英,等.基于脑电信号的疼痛强度识别方法研究[J].中国医学物理学杂志,2019,36(7):836.[doi:DOI:10.3969/j.issn.1005-202X.2019.07.017]
　LI Dong,JIN Tao,FENG Zhiying,et al.Pain intensity recognition based on EEG signals[J].Chinese Journal of Medical Physics,2019,36(6):836.[doi:DOI:10.3969/j.issn.1005-202X.2019.07.017]
[11]崔招焕,等.大鼠癫痫脑电信号采集[J].中国医学物理学杂志,2016,33(2):118.[doi:10.3969/j.issn.1005-202X.2016.02.003]
　[J].Chinese Journal of Medical Physics,2016,33(6):118.[doi:10.3969/j.issn.1005-202X.2016.02.003]
[12]张宣,刘安康,张培玲.基于嵌入式AI的癫痫发作监测系统实现[J].中国医学物理学杂志,2022,39(9):1151.[doi:DOI:10.3969/j.issn.1005-202X.2022.09.016]
　ZHANG Xuan,LIU Ankang,ZHANG Peiling.Implementation of seizure monitoring system based on embedded AI[J].Chinese Journal of Medical Physics,2022,39(6):1151.[doi:DOI:10.3969/j.issn.1005-202X.2022.09.016]
[13]汤秣雄,李效龙.一种无线闭环迷走神经刺激器及系统[J].中国医学物理学杂志,2022,39(12):1530.[doi:DOI:10.3969/j.issn.1005-202X.2022.12.012]
　TANG Moxiong,LI Xiaolong.Wireless closed-loop vagus nerve stimulator and its system[J].Chinese Journal of Medical Physics,2022,39(6):1530.[doi:DOI:10.3969/j.issn.1005-202X.2022.12.012]
[14]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(6):381.[doi:10.3969/j.issn.1005-202X.2025.03.015]

备注/Memo

备注/Memo:
【收稿日期】2025-12-16 【基金项目】国家自然科学基金（52274205）；国家重点研发计划（2018YFB1403303） 【作者简介】于洪仕，博士，讲师，研究方向：复杂网络和深度学习，E-mail: yuhongshi@lntu.edu.cn
更新日期/Last Update: 2026-06-29