EEG-Based Subjects Identification Based on Biometrics of Imagined Speech Using EMD

Abstract

When brain activity is translated into commands for real applications, the potential for human capacities augmentation is promising. In this paper, EMD is used to decompose EEG signals during Imagined Speech in order to use it as a biometric marker for creating a Biometric Recognition System. For each EEG channel, the most relevant Intrinsic Mode Functions (IMFs) are decided based on the Minkowski distance, and for each IMF 4 features are computed: Instantaneous and Teager energy distribution and Higuchi and Petrosian Fractal Dimension. To test the proposed method, a dataset with 20 subjects who imagined 30 repetitions of 5 words in Spanish, is used. Four classifiers are used for this task -random forest, SVM, naive Bayes and k-NN -and their performances are compared. The accuracy obtained (up to 0.92 using Linear SVM) after 10-folds cross-validation suggest that the proposed method based on EMD can be valuable for creating EEG-based biometrics of imagined speech for Subjects identification.

Multi-objective optimization for EEG channel selection and accurate intruder detection in an EEG-based subject identification system

“…In our previous works, we have studied and compared various paradigms, i.e. resting-state potentials and ERPs, using various types of electrodes, a various number of channels, and channel localization [2][3][4]6 . Several parameters are yet to be optimized.…”

“…A number of different paradigms are used to stimulate and record EEG signals in various subject identification approaches: imagined speech [1][2][3] , resting-state potentials 4,5 , and event-related potentials (ERPs) 6 . ERPs are very low-voltage signals emitted by the human brain that appears on the scalp in response to specific events or stimuli.…”

“…An important element is dimensionality reduction, which can be tackled through channel selection and feature extraction. Several approaches can be used to accomplish this task, including those based on methods such as principal component analysis (PCA) 13 , discrete wavelet transform (DWT) 3 , empirical mode decomposition (EMD) 2,3,6 , and even approaches using raw data as input for different configurations of neural networks (NN) 14,15 . Based on the current state-of-the-art and the results of our previous studies, we used EMD for sub-band extraction and then extracted four features for each: instantaneous and Teager energy and Higuchi and Petrosian fractal dimensions.…”

Analysis of the Precipitation Climate Signal Using Empirical Mode Decomposition (Emd) Over the Caspian Catchment Area

“…Where xi and yi are the i-th respective samples of the observed signal and the extracted IMF [5] [7]. The dmink computes for the set of all extracted IMFs.…”

“…It should be noted that the redundant IMFs may increase the redundant information and therefore reduce the Signal to Noise Ratio (SNR). On the other hand, when an IMF is inappropriate, the dmink is a maximum value due to differences in shape and frequency content rather than the original signal [5] [7]. The second criterion is the correlation coefficient that as below equation [5][12]:…”