ECoG classification based on wavelet variance.
- Author:
Shiyu YAN
1
;
Chong LIU
;
Hong WANG
;
Haibin ZHAO
Author Information
1. School of Mechanical Engineering & Automation, Northeastern University, Shenyang 110004, China. shyyan@me.neu.edu.cn
- Publication Type:Journal Article
- MeSH:
Algorithms;
Brain-Computer Interfaces;
Cerebral Cortex;
physiology;
Electroencephalography;
methods;
Humans;
Signal Processing, Computer-Assisted;
Wavelet Analysis
- From:
Journal of Biomedical Engineering
2013;30(3):460-463
- CountryChina
- Language:Chinese
-
Abstract:
For a typical electrocorticogram (ECoG)-based brain-computer interface (BCI) system in which the subject's task is to imagine movements of either the left small finger or the tongue, we proposed a feature extraction algorithm using wavelet variance. Firstly the definition and significance of wavelet variance were brought out and taken as feature based on the discussion of wavelet transform. Six channels with most distinctive features were selected from 64 channels for analysis. Consequently the EEG data were decomposed using db4 wavelet. The wavelet coeffi-cient variances containing Mu rhythm and Beta rhythm were taken out as features based on ERD/ERS phenomenon. The features were classified linearly with an algorithm of cross validation. The results of off-line analysis showed that high classification accuracies of 90. 24% and 93. 77% for training and test data set were achieved, the wavelet vari-ance had characteristics of simplicity and effectiveness and it was suitable for feature extraction in BCI research. K
