Epileptic EEG Classification by Using Time-Frequency Images for Deep Learning

2021 ◽  
pp. 2150026
Author(s):  
Mehmet Akif Ozdemir ◽  
Ozlem Karabiber Cura ◽  
Aydin Akan
Keyword(s):  
Epileptic Seizures ◽  
Eeg Signals ◽  
Clinical Tool ◽  
Energy Distributions ◽  
Time Frequency ◽  
Synchrosqueezing Transform ◽  
Novel Method ◽  
Eeg Recordings ◽  
Accuracy Rates ◽  
Aided Diagnosis

Epilepsy is one of the most common brain disorders worldwide. The most frequently used clinical tool to detect epileptic events and monitor epilepsy patients is the EEG recordings. There have been proposed many computer-aided diagnosis systems using EEG signals for the detection and prediction of seizures. In this study, a novel method based on Fourier-based Synchrosqueezing Transform (SST), which is a high-resolution time-frequency (TF) representation, and Convolutional Neural Network (CNN) is proposed to detect and predict seizure segments. SST is based on the reassignment of signal components in the TF plane which provides highly localized TF energy distributions. Epileptic seizures cause sudden energy discharges which are well represented in the TF plane by using the SST method. The proposed SST-based CNN method is evaluated using the IKCU dataset we collected, and the publicly available CHB-MIT dataset. Experimental results demonstrate that the proposed approach yields high average segment-based seizure detection precision and accuracy rates for both datasets (IKCU: 98.99% PRE and 99.06% ACC; CHB-MIT: 99.81% PRE and 99.63% ACC). Additionally, SST-based CNN approach provides significantly higher segment-based seizure prediction performance with 98.54% PRE and 97.92% ACC than similar approaches presented in the literature using the CHB-MIT dataset.

Emotion recognition using time–frequency ridges of EEG signals based on multivariate synchrosqueezing transform

2021 ◽  
Vol 0 (0) ◽  
Author(s):  
Ahmet Mert ◽  
Hasan Huseyin Celik
Keyword(s):  
Emotion Recognition ◽  
Feature Vector ◽  
Dimensional Space ◽  
Computational Cost ◽  
Feature Space ◽  
Maximum Energy ◽  
Machine Learning Algorithms ◽  
Eeg Signals ◽  
Time Frequency ◽  
Synchrosqueezing Transform

Abstract The feasibility of using time–frequency (TF) ridges estimation is investigated on multi-channel electroencephalogram (EEG) signals for emotional recognition. Without decreasing accuracy rate of the valence/arousal recognition, the informative component extraction with low computational cost will be examined using multivariate ridge estimation. The advanced TF representation technique called multivariate synchrosqueezing transform (MSST) is used to obtain well-localized components of multi-channel EEG signals. Maximum-energy components in the 2D TF distribution are determined using TF-ridges estimation to extract instantaneous frequency and instantaneous amplitude, respectively. The statistical values of the estimated ridges are used as a feature vector to the inputs of machine learning algorithms. Thus, component information in multi-channel EEG signals can be captured and compressed into low dimensional space for emotion recognition. Mean and variance values of the five maximum-energy ridges in the MSST based TF distribution are adopted as feature vector. Properties of five TF-ridges in frequency and energy plane (e.g., mean frequency, frequency deviation, mean energy, and energy deviation over time) are computed to obtain 20-dimensional feature space. The proposed method is performed on the DEAP emotional EEG recordings for benchmarking, and the recognition rates are yielded up to 71.55, and 70.02% for high/low arousal, and high/low valence, respectively.

scholarly journals Analysis of Seizure EEG in Kindled Epileptic Rats

2007 ◽  
Vol 8 (4) ◽  
pp. 225-234 ◽  
Author(s):  
A. K. Sen ◽  
M. J. Kubek ◽  
H. E. Shannon
Keyword(s):  
Epileptic Seizures ◽  
Wavelet Spectrum ◽  
Eeg Signal ◽  
Eeg Signals ◽  
Time Frequency ◽  
Frequency Representation ◽  
Frequency Changes ◽  
Global Wavelet Spectrum ◽  
Frequency Variations ◽  
Dominant Frequencies

Using wavelet analysis we have detected the presence of chirps in seizure EEG signals recorded from kindled epileptic rats. Seizures were induced by electrical stimulation of the amygdala and the EEG signals recorded from the amygdala were analyzed using a continuous wavelet transform. A time–frequency representation of the wavelet power spectrum revealed that during seizure the EEG signal is characterized by a chirp-like waveform whose frequency changes with time from the onset of seizure to its completion. Similar chirp-like time–frequency profiles have been observed in newborn and adult patients undergoing epileptic seizures. The global wavelet spectrum depicting the variation of power with frequency showed two dominant frequencies with the largest amounts of power during seizure. Our results indicate that a kindling paradigm in rats can be used as an animal model of human temporal lobe epilepsy to detect seizures by identifying chirp-like time–frequency variations in the EEG signal.

DECOMPOSITION OF SEIZURES IN EEG SIGNALS USING DISCRETE WAVELET TRANSFORMS (DWT)

2021 ◽  
pp. 50-52
Author(s):  
N Shweta ◽  
Nagendra H
Keyword(s):  
Neurological Disorders ◽  
Wavelet Transforms ◽  
Epileptic Seizures ◽  
Discrete Wavelet ◽  
Discrete Wavelet Transforms ◽  
Eeg Signals ◽  
Seizure Duration ◽  
Eeg Recordings ◽  
Electroencephalogram Eeg ◽  
The Brain

An electroencephalogram (EEG) is a test that records electrical activity in the brain. Epileptic seizures affect approximately 50 million people worldwide, making it one of the most serious neurological disorders. Seizures cause a loss of consciousness, but there are no specic signs associated with epileptic seizures. analysing the brain's activity during seizures and locating the seizure duration in EEG recordings is difcult and time consuming. A discrete wavelet transform (DWT), which is an effective tool for decomposing EEG signals into delta, theta, alpha, beta, and gamma ( and ) frequency bands. For research, the db4 is used, which has a morphological d,q,a,b g structure that is different to that of EEG.

Classification of Psychogenic Non-epileptic Seizures Using Synchrosqueezing Transform of EEG Signals

2021 ◽  
Author(s):  
Ozlem Karabiber Cura ◽  
Gulce Cosku Yilmaz ◽  
Hatice Sabiha Ture ◽  
Aydin Akan
Keyword(s):  
Epileptic Seizures ◽  
Eeg Signals ◽  
Synchrosqueezing Transform

scholarly journals Determinant of Covariance Matrix Model Coupled with AdaBoost Classification Algorithm for EEG Seizure Detection

Diagnostics ◽  
2021 ◽  
Vol 12 (1) ◽  
pp. 74
Author(s):  
Shahab Abdulla ◽  
Mohammed Diykh ◽  
Sarmad K. D. Alkhafaji ◽  
Jonathan H. Greena ◽  
Hanan Al-Hadeethi ◽  
...  
Keyword(s):  
Covariance Matrix ◽  
Back Propagation ◽  
Epileptic Seizures ◽  
Back Propagation Neural Network ◽  
Machine Learning Techniques ◽  
Eeg Signals ◽  
Average Accuracy ◽  
Effective Selection ◽  
Kolmogorov Smirnov ◽  
Eeg Recordings

Experts usually inspect electroencephalogram (EEG) recordings page-by-page in order to identify epileptic seizures, which leads to heavy workloads and is time consuming. However, the efficient extraction and effective selection of informative EEG features is crucial in assisting clinicians to diagnose epilepsy accurately. In this paper, a determinant of covariance matrix (Cov–Det) model is suggested for reducing EEG dimensionality. First, EEG signals are segmented into intervals using a sliding window technique. Then, Cov–Det is applied to each interval. To construct a features vector, a set of statistical features are extracted from each interval. To eliminate redundant features, the Kolmogorov–Smirnov (KST) and Mann–Whitney U (MWUT) tests are integrated, the extracted features ranked based on KST and MWUT metrics, and arithmetic operators are adopted to construe the most pertinent classified features for each pair in the EEG signal group. The selected features are then fed into the proposed AdaBoost Back-Propagation neural network (AB_BP_NN) to effectively classify EEG signals into seizure and free seizure segments. Finally, the AB_BP_NN is compared with several classical machine learning techniques; the results demonstrate that the proposed mode of AB_BP_NN provides insignificant false positive rates, simpler design, and robustness in classifying epileptic signals. Two datasets, the Bern–Barcelona and Bonn datasets, are used for performance evaluation. The proposed technique achieved an average accuracy of 100% and 98.86%, respectively, for the Bern–Barcelona and Bonn datasets, which is considered a noteworthy improvement compared to the current state-of-the-art methods.

scholarly journals Feature Extraction and Enhanced Convolutional Neural Network (ECNN) for Detection and Diagnosis of Seizure using EEG Signals

2019 ◽  
Vol 9 (1) ◽  
pp. 1369-1376
Keyword(s):  
Neural Network ◽  
Feature Extraction ◽  
Convolutional Neural Network ◽  
Epileptic Seizures ◽  
Classification Performance ◽  
Experimental Result ◽  
Linear Filter ◽  
Least Mean Square ◽  
Eeg Signals ◽  
Time Frequency

Seizure detection in non-stationary electroencephalography (EEG) is perplexing and difficult task. The human examination for detecting the seizure activities in EEG signals is liable to errors. Apart from the errors, it is a time driven task and also the detection is not precise. In order to detect epileptic seizures more precisely various automatic systems have been emerged to assist neurophysiologists by researchers in various attempts. There are various limitations such as time-consuming, technical artifact issues, result variation with respect to reader expertise level, abnormalities identification. Enhanced Convolutional Neural Network (ECNN) is a technique proposed to mitigate the above mentioned limitations and to categorize more accurate epileptic seizures results. A novel automatic method to sense epileptic seizures using feature extraction and detection is proposed in this research. Linear filter is helpful in reducing the noise along with artifacts when the EEG signals are preprocessed. The noise can be still removed by applying Least Mean Square algorithm. In this proposed research the features are extracted via analytic time frequency with Cascaded wavelet transform and fractal dimension (FD) in order to detect epileptic seizures. Lastly, to analyze the EEG signal for better classification performance of the given dataset, ECNN is adopted. During this research to classify normal, preictal, and seizure classes, a 13-layer deep ECNN algorithm is implemented. This research has special characteristics such that the model yields promising classification accuracy. The experimental result demonstrates that the proposed ECNN is superior in terms of higher sensitivity, specificity, accuracy and lower time complexity rather than the existing methods.

Electroencephalography and evoked potentials: a PC-based analysis program for laboratory courses in physiology.

1992 ◽  
Vol 263 (6) ◽  
pp. S16
Author(s):  
M Illert ◽  
H Wiese ◽  
U Wolfram
Keyword(s):  
Evoked Potentials ◽  
Epileptic Seizures ◽  
Dental Students ◽  
Sleep Stages ◽  
Data Sets ◽  
Eeg Signals ◽  
Cerebral Activity ◽  
Laboratory Courses ◽  
Mental Activation ◽  
Eeg Recordings

A computer program (EEG Analysis) was developed for the preclinical laboratory course in physiology held for medical and dental students. It offers an off-line analysis of a set of typical and frequently occurring physiological and pathological electroencephalogram (EEG) and evoked potential (EP) recordings, which are stored in an IBM-compatible personal computer (PC) system. The users are requested to measure and analyze the data sets and to work through a base of questions relevant in the frame of the particular topic. The program is structured in several exercises: calibration, pickup of non-EEG signals (eye movements, chewing), waveforms in EEG recordings from awake subjects (alpha-waves, beta-waves), desynchronization of cerebral activity (visual activation, acoustic activation, mental activation), habituation of cerebral activity upon acoustic stimuli, EEG recordings from asleep subjects (different sleep stages, sleep-specific EEG signals), epileptic seizures, and EPs (principle of averaging, visually evoked potentials in different cortical areas). The program runs under MS-DOS and is network capable. The software structure ensures maximal flexibility for rapid changes and adaptations of the program to specific needs of a particular EEG course. The program has been used for three years, and the response from > 800 students has been consistently positive.

Deep Learning Analysis for Estimating Sleep Syndrome Detection Utilizing the Twin Convolutional Model FTC2

2021 ◽  
Vol 1 (1) ◽  
pp. 11-17
Author(s):  
Tim Cvetko ◽  
◽  
Tinkara Robek ◽  
Keyword(s):  
Sleep Stage ◽  
Daily Basis ◽  
Sleep Stages ◽  
Time Frequency ◽  
Novel Approach ◽  
Eeg Data ◽  
Stage Classification ◽  
Sleep Stage Classification ◽  
Novel Method ◽  
Eeg Recordings

Sleep specialists often conduct manual sleep stage scoring by visually inspecting the patient's neurophysiological signals collected at sleep labs. This is a difficult, tedious and a time-consuming task. The limitations of manual sleep stage scor- ing have escalated the demand for developing Automatic Sleep Stage Classification (ASSC) systems. Sleep stage classification refers to identifying the various stages of sleep and is a critical step in an effort to assist physicians in the diag- nosis and treatment of related sleep disorders. In this paper, we propose a novel method and a practical approach to predicting early onsets of sleep syndromes utilizing the Twin Convolutional Model FTC2, including restless leg syndrome, insomnia, based on an algorithm which is comprised of two modules. A Fast Fourier Transform is applied to 30 seconds long epochs of EEG recordings to provide localized time-frequency information, and a deep convolutional LSTM neural network is trained for sleep stage classification. Automating sleep stages detection from EEG data offers a great potential to tackling sleep irregularities on a daily basis. Thereby, a novel approach for sleep stage classification is pro- posed which combines the best of signal processing and statistics. In this study, we used the PhysioNet Sleep European Data Format (EDF) Database. The code evaluation showed impressive results, reaching accuracy of 90.43, precision of 77.76, recall of 93,32, F1-score of 89.12 with the final mean false error loss 0.09. All the source code is availlable at https://github.com/timothy102/eeg.

Sign in / Sign up

Export Citation Format

Share Document