Classification of Motor Imagery BCI Using Multivariate Empirical Mode Decomposition

Abstract

Abstract-Brain electrical activity recorded via electroencephalogram (EEG) is the most convenient means for brain-computer interface (BCI), and is notoriously noisy. The information of interest is located in well defined frequency bands, and a number of standard frequency estimation algorithms have been used for feature extraction. To deal with data nonstationarity, low signal-to-noise ratio, and closely spaced frequency bands of interest, we investigate the effectiveness of recently introduced multivariate extensions of empirical mode decomposition (MEMD) in motor imagery BCI. We show that direct multichannel processing via MEMD allows for enhanced localization of the frequency information in EEG, and, in particular, its noise-assisted mode of operation (NA-MEMD) provides a highly localized time-frequency representation. Comparative analysis with other state of the art methods on both synthetic benchmark examples and a well established BCI motor imagery dataset support the analysis.Index Terms-Brain-computer interface (BCI), electroencephalogram (EEG), empirical mode decomposition, motor imagery paradigm, noise assisted multivariate extensions of empirical mode decomposition (NA-MEMD).

Estimation of Brain Connectivity during Motor Imagery Tasks using Noise-Assisted Multivariate Empirical Mode Decomposition

“…EEMD defines the final IMF as the mean of an ensemble of IMFs produced by decomposing the signal added by white noise of finite amplitude. However, the uniqueness problem is not fully resolved by EEMD and is limited by its computational and univariate nature [5]. Rehman and Mandic [14] solved these problems in their development of MEMD, which is a natural extension of the original EMD.…”

“…MIbased BCI systems can be a novel interaction option for those with motor disabilities because they do not require voluntary muscle control [4]. The physiological basis for such an MI paradigm is the mu (8)(9)(10)(11)(12) and beta rhythms (18)(19)(20)(21)(22)(23)(24)(25) in the EEG, which are found in the motor cortex region of the brain when subjects imagine movement of their hands or fingers [5]. Currently, bandpass filters, such as infinite impulse response (IIR) filters, are often used to extract the power features in the frequency bands relevant to MI tasks [1].…”

“…Recently developed multivariate empirical mode decomposition (MEMD) approach can circumvent the uniqueness problem and produce more accurate time-frequency decomposition of the intrinsic oscillations in multichannel data than original EMD, when monocomponent oscillations of interest are shared by multiple channels [14]. Several studies have investigated the functional connectivity that occurs during MI tasks to demonstrate the ability of the connectivity feature to improve the performance of the MI BCI system [4,5]. Furthermore, the EMD algorithm has been used to improve the accuracy of brain connectivity information during mental tasks [8,15].…”

Time-Frequency Analysis of Electrohysterogram for Classification of Term and Preterm Birth

“…MEMD is a time-frequency analysis technique that is optimal for nonlinear and nonstationary signals. MEMD is fully data-driven without sinusoidal basis functions, whose algorithm is described in Table 1 and [4].…”

“…Therefore, this paper proposes a novel preterm birth prediction method that applies multivariate empirical mode decomposition (MEMD) to prefilter EHG. MEMD is a time-frequency analysis technique that has been proven to be optimal for analyzing nonlinear and nonstationary signals [4]. The proposed method extracts sample entropy (SampEn) feature from the prefiltered data and classifies the term and preterm group with SampEn features.…”