Search

Sleep Apnea is a sleep disorder which causes stop in breathing for a short duration of time that happens to human beings and animals during sleep. Electroencephalogram (EEG) plays a vital role in detecting the sleep apnea by sensing and recording the brain's activities. The EEG signal dataset is subjected to filtering by using Infinite Impulse Response Butterworth Band Pass Filter and Hilbert Huang Transform. After pre-processing, the filtered EEG signal is manipulated for sub-band separation and it is fissioned into five frequency bands such as Gamma, Beta, Alpha, Theta, and Delta. This work employs features such as energy, entropy, and variance which are computed for each frequency band obtained from the decomposed EEG signals. The selected features are imported for the classification process by using machine learning classifiers including Support Vector Machine (SVM) with Kernel Functions, K-Nearest Neighbors (KNN), and Artificial Neural Network (ANN). The performance measures such as accuracy, sensitivity, and specificity are computed and analyzed for each classifier and it is inferred that the Support Vector Machine based classification of sleep apnea produces promising results.

Background: Phase-amplitude coupling (PAC)--the dependence of the amplitude of one rhythm on the phase of another, lower-frequency rhythm - has recently been used to illuminate cross-frequency coordination in neurophysiological activity. An essential step in measuring PAC is decomposing data to obtain rhythmic components of interest. Current methods of PAC assessment employ narrowband Fourier-based filters, which assume that biological rhythms are stationary, harmonic oscillations. However, biological signals frequently contain irregular and nonstationary features, which may contaminate rhythms of interest and complicate comodulogram interpretation, especially when frequency resolution is limited by short data segments., New Method: To better account for nonstationarities while maintaining sharp frequency resolution in PAC measurement, even for short data segments, we introduce a new method of PAC assessment which utilizes adaptive and more generally broadband decomposition techniques - such as the empirical mode decomposition (EMD). To obtain high frequency resolution PAC measurements, our method distributes the PAC associated with pairs of broadband oscillations over frequency space according to the time-local frequencies of these oscillations., Comparison With Existing Methods: We compare our novel adaptive approach to a narrowband comodulogram approach on a variety of simulated signals of short duration, studying systematically how different types of nonstationarities affect these methods, as well as on EEG data., Conclusions: Our results show: (1) narrowband filtering can lead to poor PAC frequency resolution, and inaccuracy and false negatives in PAC assessment; (2) our adaptive approach attains better PAC frequency resolution and is more resistant to nonstationarities and artifacts than traditional comodulograms., (Copyright © 2014 Elsevier B.V. All rights reserved.)

The communication between the human brain and the external devices can be established using Electroencephalograms (EEG)-based Brain–Computer Interface by converting the neural activities of the brain into electric signals. The EEG signals were isolated into an energy–frequency–time spectrum with Hilbert Huang transform that was used by the Deep Learning (DL)-based model to learn discriminative spectro-temporal patterns of the raw EEG signals of ten digits. This paper has two major contributions: first, create a novel dataset known as BrainDigiData of EEG signals of ten digits from (0–9) using a multi-channel EEG device. Second to propose a DL-based one-dimensional Convolutional neural network model BrainDigiCNN to classify the BrainDigiData of EEG signals of digits. The publicly available Mind Big Dataset (MBD) of digits was also used to evaluate the performance of the proposed model. The research done in this paper showed that the band-wise analysis of EEG signals in a complex scenario resulted in improved results as compared to the scenario used in the previously existing work for digit classification using EEG signals. The proposed BrainDigiCNN model achieved the highest average accuracy of 96.99%. The average classification accuracy of 98.27% was achieved for the MBD dataset of 14 channel device EMOTIV EPOC+ and 89.62% on the MBD dataset of 5-channel EMOTIV Insight. The statistical analysis of the proposed model on traditional Machine Learning (ML) classifiers using paired t-test resulted in a p-value less than 0.05 which shows the significant difference between the proposed model and ML classifiers. [ABSTRACT FROM AUTHOR]

Psychotic disorders are mental disorders that negatively affect human life. Diagnosis of psychotic patients is usually done in consultation with the patient, and this is a time-consuming process. In this study, a Computer Aided Diagnosis (CAD) system that will support expert opinion with automatic diagnosis of schizophrenia (SZ) disease, which is the leading psychotic disorder, is presented. In this study, Hilbert Huang Transform (HHT) method was used to analyze the non-stationary and non-periodic structure of EEG (Electroencephalograph) signals in the best way. The data set we used in our study includes 19-channel EEG signals from 28 (14 SZ and 14 healthy controls) participants, and the second data set includes 16-channel EEG signals from 84 (45 SZ and 39 healthy controls) participants. First of all, HS (Hilbert Spectrum) images of the first four Intrinsic Mode Functions (IMF) components obtained by applying Empirical Mode Decomposition (EMD) to EEG signals were created. These images were then classified with the VGG16 pre-trained Convolutional Neural Network (CNN) network. With our proposed method, the highest classification performance was obtained as 98.2% for Dataset I and 96.02% for Dataset II, respectively, by training the HS images obtained from the IMF 1 component with the VGG16 pre-trained CNN network. In the next step, classification performances were tested with VGG16, XCeption, DenseNet121, ResNet152 and Inception V3 pre-trained CNN networks. The high classification success achieved by the proposed method in our study demonstrates the accuracy of the model in distinguishing between SZ and healthy control. [ABSTRACT FROM AUTHOR]

The time-varying cross-spectrum method has been used to effectively study transient and dynamic brain functional connectivity between non-stationary electroencephalography (EEG) signals. Wavelet-based cross-spectrum is one of the most widely implemented methods, but it is limited by the spectral leakage caused by the finite length of the basic function that impacts the time and frequency resolutions. This paper proposes a new time-frequency brain functional connectivity analysis framework to track the non-stationary association of two EEG signals based on a Revised Hilbert-Huang Transform (RHHT). The framework can estimate the cross-spectrum of decomposed components of EEG, followed by a surrogate significance test. The results of two simulation examples demonstrate that, within a certain statistical confidence level, the proposed framework outperforms the wavelet-based method in terms of accuracy and time-frequency resolution. A case study on classifying epileptic patients and healthy controls using interictal seizure-free EEG data is also presented. The result suggests that the proposed method has the potential to better differentiate these two groups benefiting from the enhanced measure of dynamic time-frequency association. [ABSTRACT FROM AUTHOR]

Nonconvulsive status epilepticus is a condition where the patient is exposed to abnormally prolonged epileptic seizures without evident physical symptoms. Since these continuous seizures may cause permanent brain damage, it constitutes a medical emergency. This paper proposes a method to detect nonconvulsive seizures for a further nonconvulsive status epilepticus diagnosis. To differentiate between the normal and seizure electroencephalogram (EEG), a K-Nearest Neighbor, a Radial Basis Support Vector Machine, and a Linear Discriminant Analysis classifier are used. The classifier features are obtained from the Canonical Polyadic Decomposition (CPD) and Block Term Decomposition of the EEG data represented as third order tensor. To expand the EEG into a tensor, Wavelet or Hilbert-Huang transform are used. The algorithm is tested on a scalp EEG database of 139 seizures of different duration. The experimental results suggest that a Hilbert-Huang tensor representation and the CPD analysis provide the most suitable framework for nonconvulsive seizure detection. The Radial Basis Support Vector Machine classifier shows the best performance with sensitivity, specificity, and accuracy values over 98%. A rough comparison with other methods proposed in the literature shows the superior performance of the proposed method for nonconvulsive epileptic seizure detection. [ABSTRACT FROM AUTHOR]

The development of portable and wireless instruments that measure the electrical activity of the cerebral cortex (EEG) allows the capture and analysis of neurobiological signals in multiple applications. A lot of neurological, psychological and psychiatric disorders have been evaluated by EEG. Traditional EEG data-analysis methods consider linear data and stationary processes. In particular, Fourier transform deals with signals that are composed of superimposed sinusoidal oscillations, and are signals of constant frequency and amplitude. This study analyzes non-linear and non-stationary data processing using Hilbert Huang Transform (HHT), and proposes a new topography method that allows representing the brain activity. The Hilbert Transform performed on each IMF component, allows transforming the spatio-temporal data to time-frequency space, computing the amplitude and instantaneous frequency for every IMF at every time-step. An initial taxonomy concluded among pairs of channels, where IMF defines a form factor, and the pair (A, f) defines the gap between the compared signals. [ABSTRACT FROM AUTHOR]

The electroencephalogram (EEG) is an important technique that allows people to study brain signals. Classifying whether a person is concentrating or not is almost impossible with our naked eye. Thus, a system that could accurately distinguish these two categories is highly valued. Previously, the study to determine a mental state of concentration and non-concentration is based on the Fourier Transform. In this study, we employed Hilbert Huang Transform (HHT) to extract important features for this mental classification task. HHT consists of two step: (1) Using EMD and (2) Hilbert Transform. The EMD will decompose the signal into a collection of Intrinsic Mode Functions (IMF). Then Hilbert Transform is employed to obtain the IMF to compute the power spectrum to extract important features based on different subbands i.e. gamma, beta, alpha, theta and delta. An Extreme Learning Machine (ELM) is then used as a classifier to distinguish the two mental states. Four different sets of investigations were performed using the existing FFT and three investigated the HHT framework including single IMF, two and three IMFs. The results showed that the HHT with single IMF offers slight improvement compared to the existing FFT. [ABSTRACT FROM AUTHOR]

Highlights: [•] A new model for monitoring the sleep stages is built based on Hilbert Huang transform. [•] Two main oscillations were defined to depict the feature of sleep EEG based on HHT. [•] Slow- and fast-waves oscillations correspond to fluctuations in the delta and high frequency band. [•] DFA α1 was used to reflect the ANS activity during sleep. [•] The relationship between sleep EEG and HRV was significantly confirmed in this study. [ABSTRACT FROM AUTHOR]