[1]郑龙鑫,苗敏敏,徐宝国,等.运动想象脑电多视角深度森林解码算法[J].中国医学物理学杂志,2022,39(9):1159-1166.[doi:DOI:10.3969/j.issn.1005-202X.2022.09.017]
 ZHENG Longxin,MIAO Minmin,XU Baoguo,et al.Multi view deep forest-based decoding algorithm for motor imagery EEG[J].Chinese Journal of Medical Physics,2022,39(9):1159-1166.[doi:DOI:10.3969/j.issn.1005-202X.2022.09.017]
点击复制

运动想象脑电多视角深度森林解码算法()
分享到:

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

卷:
39卷
期数:
2022年第9期
页码:
1159-1166
栏目:
医学信号处理与医学仪器
出版日期:
2022-11-02

文章信息/Info

Title:
Multi view deep forest-based decoding algorithm for motor imagery EEG
文章编号:
1005-202X(2022)09-1159-08
作者:
郑龙鑫1苗敏敏12徐宝国3胡文军12
1.湖州师范学院信息工程学院, 浙江 湖州 313000; 2.浙江省现代农业资源智慧管理与应用研究重点实验室, 浙江 湖州 313000;3.东南大学仪器科学与工程学院, 江苏 南京 210096
Author(s):
ZHENG Longxin1 MIAO Minmin1 2 XU Baoguo3 HU Wenjun1 2
1. School of Information Engineering, Huzhou University, Huzhou 313000, China2.Zhejiang Province Key Laboratory of Smart Management and Application of Modern Agricultural Resources, Huzhou 313000, China3.School of Instrument Science and Engineering, Southeast University, Nanjing 210096, China
关键词:
运动想象脑电信号解码算法多视角特征提取深度森林
Keywords:
Keywords: motor imagery EEG signal decoding algorithm multi view feature extraction deep forest
分类号:
R318;TP391.4
DOI:
DOI:10.3969/j.issn.1005-202X.2022.09.017
文献标志码:
A
摘要:
针对运动想象脑电信号特征提取操作繁琐及解码精度低等问题,提出一种基于多视角深度森林的运动想象脑电解码算法。首先,通过子频带滤波及时间窗口划分对原始信号进行细粒度分析,生成空时频能量特征。然后,对上述空时频能量特征分别进行稀疏选择和时序扫描得到重要的浅层能量特征及多示例先验类别特征。继而,将上述两类特征进行融合构建运动想象脑电多视角特征集。最后,利用级联森林的逐层特征变换挖掘深层次的抽象特征进行脑电解码。根据脑机接口竞赛数据和自行采集的数据进行算法测试,并与单视角特征模型、传统共空间模式方法以及深度神经网络算法进行对比。在2个脑机接口竞赛数据集和1个真实数据集上分别取得了91.4%、75.2%和70.7%的最高平均分类准确率,结果表明该文所提多视角深度森林算法具有更优的分类识别准确率。
Abstract:
Abstract: A decoding algorithm based on multi view deep forest is proposed for solving the problems of complicated feature extraction and low decoding accuracy of motor imagery EEG signal. The energy feature in spatial-temporal-frequency domains which is generated by fine-grained analysis using subband filtering and time window division are processed with sparse selection and temporal scanning for obtaining important shallow energy features and multi instance a priori category features to construct a multi view feature set. Then the hierarchical feature transformation of cascaded forests is used to mine deep level abstract features for EEG coding.The algorithm is tested on two BCI competition datasets and a self-collected dataset, and it is compared with single view feature models, traditional CSP methods and deep neural network algorithms. The proposed method achieves the highest average classification accuracy (91.4%, 75.2% and 70.7%, respectively) in 2003 BCI competition dataset Ⅲ, 2008 BCI competition dataset 2b and self-collected dataset, which suggests that the decoding algorithm based on multi view deep forest has better classification performance.

相似文献/References:

[1]吴叶兰,张跃,曹璞刚,等.节律自适应的运动想象脑电空域特征提取方法[J].中国医学物理学杂志,2023,40(10):1270.[doi:DOI:10.3969/j.issn.1005-202X.2023.10.014]
 WU Yelan,ZHANG Yue,CAO Pugang,et al.Rhythm adaption method for extracting spatial features of MI-EEG[J].Chinese Journal of Medical Physics,2023,40(9):1270.[doi:DOI:10.3969/j.issn.1005-202X.2023.10.014]
[2]廉小亲,蔡沫豪,高超,等.基于微分熵及卷积神经网络的脑电运动想象分类识别[J].中国医学物理学杂志,2024,41(3):375.[doi:DOI:10.3969/j.issn.1005-202X.2024.03.016]
 LIAN Xiaoqin,CAI Mohao,et al.Motor imagery EEG classification and recognition based on differential entropy and convolutional neural network[J].Chinese Journal of Medical Physics,2024,41(9):375.[doi:DOI:10.3969/j.issn.1005-202X.2024.03.016]

备注/Memo

备注/Memo:
【收稿日期】2022-03-06 【项目基金】国家自然科学基金(62101189,U20A20228,61772198);江苏省前沿引领技术基础研究专项(BK20192004);浙江省基础公益研究项目(LGN18F020002) 【作者简介】郑龙鑫,硕士研究生,研究方向:机器学习及其在脑信息解码中的应用,E-mail: 1244542799@qq.com 【通信作者】苗敏敏,博士,硕士生导师,研究方向:生物医学信号处理、机器学习及其在脑信息解码中的应用,E-mail: 02746@zjhu.edu.cn
更新日期/Last Update: 2022-09-27