GPGPU-Aided Ensemble Empirical-Mode Decomposition for EEG Analysis During Anesthesia

Abstract

Ensemble empirical-mode decomposition (EEMD) is a novel adaptive time-frequency analysis method, which is particularly suitable for extracting useful information from noisy nonlinear or nonstationary data. Unfortunately, since the EEMD is highly compute-intensive, the method does not apply in real-time applications on top of commercial-off-the-shelf computers. Aiming at this problem, a parallelized EEMD method has been developed using general-purpose computing on the graphics processing unit (GPGPU), namely, G-EEMD. A spectral entropy facilitated by G-EEMD was, therefore, proposed to analyze the EEG data for estimating the depth of anesthesia (DoA) in a real-time manner. In terms of EEG data analysis, G-EEMD has dramatically improved the run-time performance by more than 140 times compared to the original serial EEMD implementation. G-EEMD also performs far better than another parallelized implementation of EEMD bases on conventional CPU-based distributed computing technology despite the latter utilizes 16 high-end computing nodes for the same computing task. Furthermore, the results obtained from a pharmacokinetics/pharmacodynamic (PK/PD) model analysis indicate that the EEMD method is slightly more effective than its precedent alternative method (EMD) in estimating DoA, the coefficient of determination R(2) by EEMD is significantly higher than that by EMD (p < 0.05, paired t-test) and the prediction probability P(k) by EEMD is also slighter higher than that by EMD (p < 0.2, paired t-test).

EOG Artifact Correction from EEG Recording Using Stationary Subspace Analysis and Empirical Mode Decomposition

“…Thereby it may lead to loss of cerebral information by simply rejecting the artifactual components during the reconstruction. On the other hand, despite the fact that EMD adaptively derives the decomposition basis from local dynamics in the data, it has been reported that EMD may fail to correct the artifacts by directly working on the raw EEG recordings [42]. This is because the ratio of artifact magnitude to the cerebral activity magnitude in a channel may not be high enough to separate them effectively [42].…”

Comparison between traditional FFT and marginal spectra using the Hilbert-Huang transform method for the broadband spectral analysis of the EEG in healthy humans

“…Even though the HMS has been used by other authors for particular applications on EEG spectral analysis (Chen et al 2010;Chen et al 2016;Li 2006;Li et al 2008;Xiangjun et al 2017;Zhu et al 2015;Park et al 2011) the possible differences with the traditional FFT spectrum have not been analyzed in details particularly in healthy humans, as we have described in this study, and could be considered the first report about this particular issue. The fact that we have shown no substantial differences for the typical indices calculated for broadband EEG spectral analysis using both methods in this study, can be considered a good evidence that the use of the classic FFT method, applied extensively and for many years, can be used in the conditions in which was carried out this investigation, it is, in healthy subjects, during a functional state of relaxed wakefulness during resting eyes-closed condition, and for segments of consecutive EEG of 60 seconds, but for other physiological conditions, during cognitive studies, or to study patients with different pathologies that can affect the processes responsible of the generation of the bioelectric activity of the brain, the presence of nonlinearity and particularly of non-stationarity may produce a misleading result while using the FFT methods (Alegre-Cortes et al 2016;Huang et al 1998;Munoz-Gutierrez et al 2018;Soler et al 2020;Tsai et al 2016).…”

“…The HMS has been used by different authors after applying the Hilbert-Huang method to the EEG signal (Chen et al 2010;Li 2006;Li et al 2008;Chen et al 2016;Xiangjun et al 2017;Zhu et al 2015;Park et al 2011) but to our knowledge, the possible differences between the results obtained using the HMS and the conventional FFT spectra in resting conditions in healthy humans have not been studied in detail, and it results imperative, because for many years the universally used method for the spectral analysis of the EEG has used the FFT spectra. The advantages and possible limitations of the novel alternative approach, in this case the use of the HMS, could be better appreciated and could contribute to their more extended use in EEG investigations.…”

A Novel GPU-Based Acceleration Algorithm for a Longwave Radiative Transfer Model

“…Each grid consists of thread blocks. Inside each thread block, data is efficiently shared by a group of threads through a fast shared memory [34]. CUDA supports many high-level languages, such as C/C++ and Fortran.…”