Classification of High Frequency Oscillations in intracranial EEG signals based on coupled time-frequency and image-related features

2022 ◽  
Vol 73 ◽  
pp. 103418
Author(s):  
Fatma Krikid ◽  
Ahmad Karfoul ◽  
Sahbi Chaibi ◽  
Amar Kachenoura ◽  
Anca Nica ◽  
...  
Keyword(s):  
High Frequency ◽  
Eeg Signals ◽  
Intracranial Eeg ◽  
Time Frequency ◽  
High Frequency Oscillations

Automatic Detection and Classification of High-Frequency Oscillations in Depth-EEG Signals

2017 ◽  
Vol 64 (9) ◽  
pp. 2230-2240 ◽  
Author(s):  
Nisrine Jrad ◽  
Amar Kachenoura ◽  
Isabelle Merlet ◽  
Fabrice Bartolomei ◽  
Anca Nica ◽  
...  
Keyword(s):  
High Frequency ◽  
Automatic Detection ◽  
Eeg Signals ◽  
High Frequency Oscillations

scholarly journals Artificial Neural Network Classification of Motor-Related EEG: An Increase in Classification Accuracy by Reducing Signal Complexity

Complexity ◽  
2018 ◽  
Vol 2018 ◽  
pp. 1-10 ◽  
Author(s):  
Vladimir A. Maksimenko ◽  
Semen A. Kurkin ◽  
Elena N. Pitsik ◽  
Vyacheslav Yu. Musatov ◽  
Anastasia E. Runnova ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Motor Imagery ◽  
High Frequency ◽  
Classification Accuracy ◽  
Eeg Signals ◽  
Pass Filter ◽  
Time Frequency ◽  
Artificial Neural

We apply artificial neural network (ANN) for recognition and classification of electroencephalographic (EEG) patterns associated with motor imagery in untrained subjects. Classification accuracy is optimized by reducing complexity of input experimental data. From multichannel EEG recorded by the set of 31 electrodes arranged according to extended international 10-10 system, we select an appropriate type of ANN which reaches 80 ± 10% accuracy for single trial classification. Then, we reduce the number of the EEG channels and obtain an appropriate recognition quality (up to 73 ± 15%) using only 8 electrodes located in frontal lobe. Finally, we analyze the time-frequency structure of EEG signals and find that motor-related features associated with left and right leg motor imagery are more pronounced in the mu (8–13 Hz) and delta (1–5 Hz) brainwaves than in the high-frequency beta brainwave (15–30 Hz). Based on the obtained results, we propose further ANN optimization by preprocessing the EEG signals with a low-pass filter with different cutoffs. We demonstrate that the filtration of high-frequency spectral components significantly enhances the classification performance (up to 90 ± 5% accuracy using 8 electrodes only). The obtained results are of particular interest for the development of brain-computer interfaces for untrained subjects.

scholarly journals Exclusion of the Possibility of “False Ripples” From Ripple Band High-Frequency Oscillations Recorded From Scalp Electroencephalogram in Children With Epilepsy

2021 ◽  
Vol 15 ◽  
Author(s):  
Katsuhiro Kobayashi ◽  
Takashi Shibata ◽  
Hiroki Tsuchiya ◽  
Tomoyuki Akiyama
Keyword(s):  
High Frequency ◽  
Numerical Differentiation ◽  
Convincing Evidence ◽  
Exact Test ◽  
Time Frequency ◽  
High Frequency Oscillations ◽  
Eeg Data ◽  
Group A ◽  
Spurious Oscillations ◽  
Scalp Electroencephalogram

AimRipple-band epileptic high-frequency oscillations (HFOs) can be recorded by scalp electroencephalography (EEG), and tend to be associated with epileptic spikes. However, there is a concern that the filtration of steep waveforms such as spikes may cause spurious oscillations or “false ripples.” We excluded such possibility from at least some ripples by EEG differentiation, which, in theory, enhances high-frequency signals and does not generate spurious oscillations or ringing.MethodsThe subjects were 50 pediatric patients, and ten consecutive spikes during sleep were selected for each patient. Five hundred spike data segments were initially reviewed by two experienced electroencephalographers using consensus to identify the presence or absence of ripples in the ordinary filtered EEG and an associated spectral blob in time-frequency analysis (Session A). These EEG data were subjected to numerical differentiation (the second derivative was denoted as EEG″). The EEG″ trace of each spike data segment was shown to two other electroencephalographers who judged independently whether there were clear ripple oscillations or uncertain ripple oscillations or an absence of oscillations (Session B).ResultsIn Session A, ripples were identified in 57 spike data segments (Group A-R), but not in the other 443 data segments (Group A-N). In Session B, both reviewers identified clear ripples (strict criterion) in 11 spike data segments, all of which were in Group A-R (p < 0.0001 by Fisher’s exact test). When the extended criterion that included clear and/or uncertain ripples was used in Session B, both reviewers identified 25 spike data segments that fulfilled the criterion: 24 of these were in Group A-R (p < 0.0001).DiscussionWe have demonstrated that real ripples over scalp spikes exist in a certain proportion of patients. Ripples that were visualized consistently using both ordinary filters and the EEG″ method should be true, but failure to clarify ripples using the EEG″ method does not mean that true ripples are absent.ConclusionThe numerical differentiation of EEG data provides convincing evidence that HFOs were detected in terms of the presence of such unusually fast oscillations over the scalp and the importance of this electrophysiological phenomenon.

EEG Analysis

2017 ◽  
Author(s):  
Fabrice Wendling ◽  
Marco Congendo ◽  
Fernando H. Lopes da Silva
Keyword(s):  
Matching Pursuit ◽  
Effective Connectivity ◽  
Eeg Signals ◽  
Model Decomposition ◽  
Time Frequency ◽  
Frequency Domains ◽  
Stationary Signals ◽  
Artifact Rejection ◽  
Transient Events

This chapter addresses the analysis and quantification of electroencephalographic (EEG) and magnetoencephalographic (MEG) signals. Topics include characteristics of these signals and practical issues such as sampling, filtering, and artifact rejection. Basic concepts of analysis in time and frequency domains are presented, with attention to non-stationary signals focusing on time-frequency signal decomposition, analytic signal and Hilbert transform, wavelet transform, matching pursuit, blind source separation and independent component analysis, canonical correlation analysis, and empirical model decomposition. The behavior of these methods in denoising EEG signals is illustrated. Concepts of functional and effective connectivity are developed with emphasis on methods to estimate causality and phase and time delays using linear and nonlinear methods. Attention is given to Granger causality and methods inspired by this concept. A concrete example is provided to show how information processing methods can be combined in the detection and classification of transient events in EEG/MEG signals.

scholarly journals Recognition of Emotional States using Multiscale Information Analysis of High Frequency EEG Oscillations

Entropy ◽  
2019 ◽  
Vol 21 (6) ◽  
pp. 609 ◽  
Author(s):  
Gao ◽  
Cui ◽  
Wan ◽  
Gu
Keyword(s):  
Emotion Recognition ◽  
High Frequency ◽  
Low Frequency ◽  
Support Vector ◽  
Discrete Wavelet ◽  
Emotional States ◽  
Circumplex Model ◽  
Information Analysis ◽  
Eeg Signals ◽  
High Frequency Oscillations

Exploring the manifestation of emotion in electroencephalogram (EEG) signals is helpful for improving the accuracy of emotion recognition. This paper introduced the novel features based on the multiscale information analysis (MIA) of EEG signals for distinguishing emotional states in four dimensions based on Russell's circumplex model. The algorithms were applied to extract features on the DEAP database, which included multiscale EEG complexity index in the time domain, and ensemble empirical mode decomposition enhanced energy and fuzzy entropy in the frequency domain. The support vector machine and cross validation method were applied to assess classification accuracy. The classification performance of MIA methods (accuracy = 62.01%, precision = 62.03%, recall/sensitivity = 60.51%, and specificity = 82.80%) was much higher than classical methods (accuracy = 43.98%, precision = 43.81%, recall/sensitivity = 41.86%, and specificity = 70.50%), which extracted features contain similar energy based on a discrete wavelet transform, fractal dimension, and sample entropy. In this study, we found that emotion recognition is more associated with high frequency oscillations (51–100Hz) of EEG signals rather than low frequency oscillations (0.3–49Hz), and the significance of the frontal and temporal regions are higher than other regions. Such information has predictive power and may provide more insights into analyzing the multiscale information of high frequency oscillations in EEG signals.

Sign in / Sign up

Export Citation Format

Share Document