Reaching the ceiling? Empirical scaling behaviour for deep EEG pathology classification

Machine learning techniques, particularly deep convolutional neural networks (ConvNets), are increasingly being used to automate clinical EEG analysis, with the potential to reduce the clinical burden and improve patient care. However, further research is required before they can be used in clinical settings, particularly regarding the impact of the number of training samples and model parameters on their testing error. To address this, we present a comprehensive study of the empirical scaling behaviour of ConvNets for EEG pathology classification. We analysed the testing error with increasing the training samples and model size for four different ConvNet architectures. The focus of our experiments is width scaling, and we have increased the number of parameters to up to 1.8 million. Our evaluation was based on two publicly available datasets: the Temple University Hospital (TUH) Abnormal EEG Corpus and the TUH Abnormal Expansion Balanced EEG Corpus, which together contain 10,707 training samples. The results show that the testing error follows a saturating power-law with both model and dataset size. This pattern is consistent across different datasets and ConvNet architectures. Furthermore, empirically observed accuracies saturate at 85%-87%, which may be due to an imperfect inter-rater agreement on the clinical labels. The empirical scaling behaviour of the test performance with dataset and model size has significant implications for deep EEG pathology classification research and practice. Our findings highlight the potential of deep ConvNets for high-performance EEG pathology classification, and the identified scaling relationships provide valuable recommendations for the advancement of automated EEG diagnostics.

Effects of gender and age on sleep EEG functional connectivity differences in subjects with mild difficulty falling asleep

Introduction

Difficulty falling asleep place an increasing burden on society. EEG-based sleep staging is fundamental to the diagnosis of sleep disorder, and the selection of features for each sleep stage is a key step in the sleep analysis. However, the differences of sleep EEG features in gender and age are not clear enough.

Methods

This study aimed to investigate the effects of age and gender on sleep EEG functional connectivity through statistical analysis of brain functional connectivity and machine learning validation. The two-overnight sleep EEG data of 78 subjects with mild difficulty falling asleep were categorized into five sleep stages using markers and segments from the "sleep-EDF" public database. First, the 78 subjects were finely grouped, and the mutual information of the six sleep EEG rhythms of δ, θ, α, β, spindle, and sawtooth wave was extracted as a functional connectivity measure. Then, one-way analysis of variance (ANOVA) was used to extract significant differences in functional connectivity of sleep rhythm waves across sleep stages with respect to age and gender. Finally, machine learning algorithms were used to investigate the effects of fine grouping of age and gender on sleep staging.

Results and discussion

The results showed that: (1) The functional connectivity of each sleep rhythm wave differed significantly across sleep stages, with delta and beta functional connectivity differing significantly across sleep stages. (2) Significant differences in functional connections among young and middle-aged groups, and among young and elderly groups, but no significant difference between middle-aged and elderly groups. (3) Female functional connectivity strength is generally higher than male at the high-frequency band of EEG, but no significant difference in the low-frequency. (4) Finer group divisions based on gender and age can indeed improve the accuracy of sleep staging, with an increase of about 3.58% by using the random forest algorithm. Our results further reveal the electrophysiological neural mechanisms of each sleep stage, and find that sleep functional connectivity differs significantly in both gender and age, providing valuable theoretical guidance for the establishment of automated sleep stage models.

Machine learning-empowered sleep staging classification using multi-modality signals

The goal is to enhance an automated sleep staging system's performance by leveraging the diverse signals captured through multi-modal polysomnography recordings. Three modalities of PSG signals, namely electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG), were considered to obtain the optimal fusions of the PSG signals, where 63 features were extracted. These include frequency-based, time-based, statistical-based, entropy-based, and non-linear-based features. We adopted the ReliefF (ReF) feature selection algorithms to find the suitable parts for each signal and superposition of PSG signals. Twelve top features were selected while correlated with the extracted feature sets' sleep stages. The selected features were fed into the AdaBoost with Random Forest (ADB + RF) classifier to validate the chosen segments and classify the sleep stages. This study's experiments were investigated by obtaining two testing schemes: epoch-wise testing and subject-wise testing. The suggested research was conducted using three publicly available datasets: ISRUC-Sleep subgroup1 (ISRUC-SG1), sleep-EDF(S-EDF), Physio bank CAP sleep database (PB-CAPSDB), and S-EDF-78 respectively. This work demonstrated that the proposed fusion strategy overestimates the common individual usage of PSG signals.

A classification-based generative approach to selective targeting of global slow oscillations during sleep

Background

Given sleep's crucial role in health and cognition, numerous sleep-based brain interventions are being developed, aiming to enhance cognitive function, particularly memory consolidation, by improving sleep. Research has shown that Transcranial Alternating Current Stimulation (tACS) during sleep can enhance memory performance, especially when used in a closed-loop (cl-tACS) mode that coordinates with sleep slow oscillations (SOs, 0.5-1.5Hz). However, sleep tACS research is characterized by mixed results across individuals, which are often attributed to individual variability.

Objective/Hypothesis

This study targets a specific type of SOs, widespread on the electrode manifold in a short delay ("global SOs"), due to their close relationship with long-term memory consolidation. We propose a model-based approach to optimize cl-tACS paradigms, targeting global SOs not only by considering their temporal properties but also their spatial profile.

Methods

We introduce selective targeting of global SOs using a classification-based approach. We first estimate the current elicited by various stimulation paradigms, and optimize parameters to match currents found in natural sleep during a global SO. Then, we employ an ensemble classifier trained on sleep data to identify effective paradigms. Finally, the best stimulation protocol is determined based on classification performance.

Results

Our study introduces a model-driven cl-tACS approach that specifically targets global SOs, with the potential to extend to other brain dynamics. This method establishes a connection between brain dynamics and stimulation optimization.

Conclusion

Our research presents a novel approach to optimize cl-tACS during sleep, with a focus on targeting global SOs. This approach holds promise for improving cl-tACS not only for global SOs but also for other physiological events, benefiting both research and clinical applications in sleep and cognition.

Differences in intestinal motility during different sleep stages based on long-term bowel sounds

BACKGROUND AND OBJECTIVES

This study focused on changes in intestinal motility during different sleep stages based on long-term bowel sounds.

METHODS

A modified higher order statistics algorithm was devised to identify the effective bowel sound segments. Next, characteristic values (CVs) were extracted from each bowel sound segment, which included 4 time-domain, 4 frequency-domain and 2 nonlinear CVs. The statistical analysis of these CVs corresponding to the different sleep stages could be used to evaluate the changes in intestinal motility during sleep.

RESULTS

A total of 6865.81 min of data were recorded from 14 participants, including both polysomnographic data and bowel sound data which were recorded simultaneously from each participant. The average accuracy, sensitivity and specificity of the modified higher order statistics detector were 96.46 ± 2.60%, 97.24 ± 2.99% and 94.13 ± 4.37%. In addition, 217088 segments of effective bowel sound corresponding to different sleep stages were identified using the modified detector. Most of the CVs were statistically different during different sleep stages ([Formula: see text]). Furthermore, the bowel sounds were low in frequency based on frequency-domain CVs, high in energy based on time-domain CVs and low in complexity base on nonlinear CVs during deep sleep, which was consistent with the state of the EEG signals during deep sleep.

CONCLUSIONS

The intestinal motility varies by different sleep stages based on long-term bowel sounds using the modified higher order statistics detector. The study indicates that the long-term bowel sounds can well reflect intestinal motility during sleep. This study also demonstrates that it is technically feasible to simultaneously record intestinal motility and sleep state throughout the night. This offers great potential for future studies investigating intestinal motility during sleep and related clinical applications.

Age-integrated artificial intelligence framework for sleep stage classification and obstructive sleep apnea screening

Introduction

Sleep is an essential function to sustain a healthy life, and sleep dysfunction can cause various physical and mental issues. In particular, obstructive sleep apnea (OSA) is one of the most common sleep disorders and, if not treated in a timely manner, OSA can lead to critical problems such as hypertension or heart disease.

Methods

The first crucial step in evaluating individuals' quality of sleep and diagnosing sleep disorders is to classify sleep stages using polysomnographic (PSG) data including electroencephalography (EEG). To date, such sleep stage scoring has been mainly performed manually via visual inspection by experts, which is not only a time-consuming and laborious process but also may yield subjective results. Therefore, we have developed a computational framework that enables automatic sleep stage classification utilizing the power spectral density (PSD) features of sleep EEG based on three different learning algorithms: support vector machine, k-nearest neighbors, and multilayer perceptron (MLP). In particular, we propose an integrated artificial intelligence (AI) framework to further inform the risk of OSA based on the characteristics in automatically scored sleep stages. Given the previous finding that the characteristics of sleep EEG differ by age group, we employed a strategy of training age-specific models (younger and older groups) and a general model and comparing their performance.

Results

The performance of the younger age-specific group model was similar to that of the general model (and even higher than the general model at certain stages), but the performance of the older age-specific group model was rather low, suggesting that bias in individual variables, such as age bias, should be considered during model training. Our integrated model yielded an accuracy of 73% in sleep stage classification and 73% in OSA screening when MLP algorithm was applied, which indicates that patients with OSA could be screened with the corresponding accuracy level only with sleep EEG without respiration-related measures.

Discussion

The current outcomes demonstrate the feasibility of AI-based computational studies that when combined with advances in wearable devices and relevant technologies could contribute to personalized medicine by not only assessing an individuals' sleep status conveniently at home but also by alerting them to the risk of sleep disorders and enabling early intervention.

Driver Drowsiness EEG Detection Based on Tree Federated Learning and Interpretable Network

Accurate identification of driver's drowsiness state through Electroencephalogram (EEG) signals can effectively reduce traffic accidents, but EEG signals are usually stored in various clients in the form of small samples. This study attempts to construct an efficient and accurate privacy-preserving drowsiness monitoring system, and proposes a fusion model based on tree Federated Learning (FL) and Convolutional Neural Network (CNN), which can not only identify and explain the driver's drowsiness state, but also integrate the information of different clients under the premise of privacy protection. Each client uses CNN with the Global Average Pooling (GAP) layer and shares model parameters. The tree FL transforms communication relationships into a graph structure, and model parameters are transmitted in parallel along connected branches of the graph. Moreover, the Class Activation Mapping (CAM) is used to find distinctive EEG features for representing specific classes. On EEG data of 11 subjects, it is found that this method has higher average accuracy, F1-score and AUC than the traditional classification method, reaching 73.56%, 73.26% and 78.23%, respectively. Compared with the traditional FL algorithm, this method better protects the driver's privacy and improves communication efficiency.

SleepXAI: An explainable deep learning approach for multi-class sleep stage identification

Extensive research has been conducted on the automatic classification of sleep stages utilizing deep neural networks and other neurophysiological markers. However, for sleep specialists to employ models as an assistive solution, it is necessary to comprehend how the models arrive at a particular outcome, necessitating the explainability of these models. This work proposes an explainable unified CNN-CRF approach (SleepXAI) for multi-class sleep stage classification designed explicitly for univariate time-series signals using modified gradient-weighted class activation mapping (Grad-CAM). The proposed approach significantly increases the overall accuracy of sleep stage classification while demonstrating the explainability of the multi-class labeling of univariate EEG signals, highlighting the parts of the signals emphasized most in predicting sleep stages. We extensively evaluated our approach to the sleep-EDF dataset, and it demonstrates the highest overall accuracy of 86.8% in identifying five sleep stage classes. More importantly, we achieved the highest accuracy when classifying the crucial sleep stage N1 with the lowest number of instances, outperforming the state-of-the-art machine learning approaches by 16.3%. These results motivate us to adopt the proposed approach in clinical practice as an aid to sleep experts.

L-Tetrolet Pattern-Based Sleep Stage Classification Model Using Balanced EEG Datasets

Background: Sleep stage classification is a crucial process for the diagnosis of sleep or sleep-related diseases. Currently, this process is based on manual electroencephalogram (EEG) analysis, which is resource-intensive and error-prone. Various machine learning models have been recommended to standardize and automate the analysis process to address these problems. Materials and methods: The well-known cyclic alternating pattern (CAP) sleep dataset is used to train and test an L-tetrolet pattern-based sleep stage classification model in this research. By using this dataset, the following three cases are created, and they are: Insomnia, Normal, and Fused cases. For each of these cases, the machine learning model is tasked with identifying six sleep stages. The model is structured in terms of feature generation, feature selection, and classification. Feature generation is established with a new L-tetrolet (Tetris letter) function and multiple pooling decomposition for level creation. We fuse ReliefF and iterative neighborhood component analysis (INCA) feature selection using a threshold value. The hybrid and iterative feature selectors are named threshold selection-based ReliefF and INCA (TSRFINCA). The selected features are classified using a cubic support vector machine. Results: The presented L-tetrolet pattern and TSRFINCA-based sleep stage classification model yield 95.43%, 91.05%, and 92.31% accuracies for Insomnia, Normal dataset, and Fused cases, respectively. Conclusion: The recommended L-tetrolet pattern and TSRFINCA-based model push the envelope of current knowledge engineering by accurately classifying sleep stages even in the presence of sleep disorders.

Automatic detection of abnormal EEG signals using multiscale features with ensemble learning

Electroencephalogram (EEG) is an economical and convenient auxiliary test to aid in the diagnosis and analysis of brain-related neurological diseases. In recent years, machine learning has shown great potential in clinical EEG abnormality detection. However, existing methods usually fail to consider the issue of feature redundancy when extracting the relevant EEG features. In addition, the importance of utilizing the patient age information in EEG detection is ignored. In this paper, a new framework is proposed for distinguishing an unknown EEG recording as either normal or abnormal by identifying different types of EEG-derived significant features. In the proposed framework, different hierarchical salient features are extracted using a time-wise multi-scale aggregation strategy, based on a selected group of statistical characteristics calculated from the optimum discrete wavelet transform coefficients. We also fuse the age information with multi-scale features for further improving discrimination. The integrated features are classified using three ensemble learning classifiers, CatBoost, LightGBM, and random forest. Experimental results show that our method with CatBoost classifier can yield superior performance vis-a-vis competing techniques, which indicates the great promise of our methodology in EEG pathology detection.

A Holistic Strategy for Classification of Sleep Stages with EEG

Manual sleep stage scoring is usually implemented with the help of sleep specialists by means of visual inspection of the neurophysiological signals of the patient. As it is a very hectic task to perform, automated sleep stage classification systems were developed in the past, and advancements are being made consistently by researchers. The various stages of sleep are identified by these automated sleep stage classification systems, and it is quite an important step to assist doctors for the diagnosis of sleep-related disorders. In this work, a holistic strategy named as clustering and dimensionality reduction with feature extraction cum selection for classification along with deep learning (CDFCD) is proposed for the classification of sleep stages with EEG signals. Though the methodology follows a similar structural flow as proposed in the past works, many advanced and novel techniques are proposed under each category in this work flow. Initially, clustering is applied with the help of hierarchical clustering, spectral clustering, and the proposed principal component analysis (PCA)-based subspace clustering. Then the dimensionality of it is reduced with the help of the proposed singular value decomposition (SVD)-based spectral algorithm and the standard variational Bayesian matrix factorization (VBMF) technique. Then the features are extracted and selected with the two novel proposed techniques, such as the sparse group lasso technique with dual-level implementation (SGL-DLI) and the ridge regression technique with limiting weight scheme (RR-LWS). Finally, the classification happens with the less explored multiclass Gaussian process classification (MGC), the proposed random arbitrary collective classification (RACC), and the deep learning technique using long short-term memory (LSTM) along with other conventional machine learning techniques. This methodology is validated on the sleep EDF database, and the results obtained with this methodology have surpassed the results of the previous studies in terms of the obtained classification accuracy reporting a high accuracy of 93.51% even for the six-classes classification problem.

A Fast Detection Method of Break Points in Effective Connectivity Networks

There is increasing interest in identifying changes in the underlying states of brain networks. The availability of large scale neuroimaging data creates a strong need to develop fast, scalable methods for detecting and localizing in time such changes and also identify their drivers, thus enabling neuroscientists to hypothesize about potential mechanisms. This paper presents a fast method for detecting break points in exceedingly long time series neurogimaging data, based on vector autoregressive (Granger causal) models. It uses a multi-step strategy based on a regularized objective function that leads to fast identification of candidate break points, followed by clustering steps to select the final set of break points and subsequent estimation with false positives control of the underlying Granger causal networks. The latter provide insights into key changes in network connectivity that led to the presence of break points. The proposed methodology is illustrated on synthetic data varying in their length, dimensionality, number of break points, strength of signal and also applied to EEG data related to visual tasks.

Quantifying Differences Between Affine and Nonlinear Spatial Normalization of FP-CIT Spect Images

Spatial normalization helps us to compare quantitatively two or more input brain scans. Although using an affine normalization approach preserves the anatomical structures, the neuroimaging field is more common to find works that make use of nonlinear transformations. The main reason is that they facilitate a voxel-wise comparison, not only when studying functional images but also when comparing MRI scans given that they fit better to a reference template. However, the amount of bias introduced by the nonlinear transformations can potentially alter the final outcome of a diagnosis especially when studying functional scans for neurological disorders like Parkinson's Disease. In this context, we have tried to quantify the bias introduced by the affine and the nonlinear spatial registration of FP-CIT SPECT volumes of healthy control subjects and patients with PD. For that purpose, we calculated the deformation fields of each participant and applied these deformation fields to a 3D-grid. As the space between the edges of small cubes comprising the grid change, we can quantify which parts from the brain have been enlarged, compressed or just remain the same. When the nonlinear approach is applied, scans from PD patients show a region near their striatum very similar in shape to that of healthy subjects. This artificially increases the interclass separation between patients with PD and healthy subjects as the local intensity is decreased in the latter region, and leads machine learning systems to biased results due to the artificial information introduced by these deformations.

A deep learning algorithm based on 1D CNN-LSTM for automatic sleep staging

BACKGROUND

Sleep staging is an important part of sleep research. Traditional automatic sleep staging based on machine learning requires extensive feature extraction and selection.

OBJECTIVE

This paper proposed a deep learning algorithm without feature extraction based on one-dimensional convolutional neural network and long short-term memory.

METHODS

The algorithm can automatically divide sleep into 5 phases including awake period, non-rapid eye movement sleep period (N1 ∼ N3) and rapid eye movement using the electroencephalogram signals. The raw signal was processed by the wavelet transform. Then, the processed signal was directly input into the deep learning algorithm to obtain the staging result.

RESULTS

The accuracy of staging is 93.47% using the Fpz-Cz electroencephalogram signal. When using the Fpz-Cz and electroencephalogram signal, the algorithm can obtain the highest accuracy of 94.15%.

CONCLUSION

These results show that this algorithm is suitable for different physiological signals and can realize end-to-end automatic sleep staging without any manual feature extraction.

Detection of k-complexes in EEG signals using a multi-domain feature extraction coupled with a least square support vector machine classifier

Sleep scoring is one of the primary tasks for the classification of sleep stages using electroencephalogram (EEG) signals. It is one of the most important diagnostic methods in sleep research and must be carried out with a high degree of accuracy because any errors in the scoring in the patient's sleep EEG recordings can cause serious problems. The aim of this research is to develop a new automatic method for detecting the most important characteristics in sleep stage 2 such as k-complexes based on multi-domain features. In this study, each EEG signal is divided into a set of segments using a sliding window technique. Based on extensive experiments during the training phase, the size of the sliding window is set to 0.5 s (s). Then a set of statistical, fractal, frequency and non-linear features are extracted from each epoch based on the time domain, Katz's algorithm, power spectrum density (PSD) and tunable Q-factor wavelet transform (TQWT). As a result, a vector of twenty-two features is obtained to represent each EEG segment. In order to detect k-complexes, the extracted features were analysed for their ability to detect the k-complex waveforms. Based on the analysis of the features, twelve out of twenty-two features are selected and forwarded to a least square support vector machine (LS-SVM) classifier to identify k-complexes in EEG signals. A set of various classification techniques of K-means and extreme learning machine classifiers are used to compare the obtained results and to evaluate the performance of the proposed method.The experimental results showed that the proposed method, based on multi-domain features, achieved better recognition results than other methods and classifiers. An average accuracy, sensitivity and specificity of 97.7 %, 97 %, and 94.2 % were obtained, respectively, with the CZ-A1 channel according to the R&K standard. The experimental results with high classification performance demonstrated that the technique can help doctors optimize the diagnosis and treatment of sleep disorders.

Quantitative investigation reveals distinct phases in Drosophila sleep

The fruit fly, Drosophila melanogaster, has been used as a model organism for the molecular and genetic dissection of sleeping behaviors. However, most previous studies were based on qualitative or semi-quantitative characterizations. Here we quantified sleep in flies. We set up an assay to continuously track the activity of flies using infrared camera, which monitored the movement of tens of flies simultaneously with high spatial and temporal resolution. We obtained accurate statistics regarding the rest and sleep patterns of single flies. Analysis of our data has revealed a general pattern of rest and sleep: the rest statistics obeyed a power law distribution and the sleep statistics obeyed an exponential distribution. Thus, a resting fly would start to move again with a probability that decreased with the time it has rested, whereas a sleeping fly would wake up with a probability independent of how long it had slept. Resting transits to sleeping at time scales of minutes. Our method allows quantitative investigations of resting and sleeping behaviors and our results provide insights for mechanisms of falling into and waking up from sleep.

A multiplex visibility graph motif‐based convolutional neural network for characterizing sleep stages using EEG signals

Sleep is an essential integrant in everyone’s daily life; therefore, it is an important but challenging problem to characterize sleep stages from electroencephalogram (EEG) signals. The network motif has been developed as a useful tool to investigate complex networks. In this study, we developed a multiplex visibility graph motif‐based convolutional neural network (CNN) for characterizing sleep stages using EEG signals and then introduced the multiplex motif entropy as the quantitative index to distinguish the six sleep stages. The independent samples t‐test shows that the multiplex motif entropy values have significant differences among the six sleep stages. Furthermore, we developed a CNN model and employed the multiplex motif sequence as the input of the model to classify the six sleep stages. Notably, the classification accuracy of the six‐state stage detection was 85.27%. Results demonstrated the effectiveness of the multiplex motif in characterizing the dynamic features underlying different sleep stages, whereby they further provide an essential strategy for future sleep‐stage detection research.

Prediction of Depression Severity Scores Based on Functional Connectivity and Complexity of the EEG Signal

BACKGROUND

Depression is one of the most common mental disorders and the leading cause of functional disabilities. This study aims to specify whether functional connectivity and complexity of brain activity can predict the severity of depression (Beck Depression Inventory-II scores).

METHODS

Resting-state, eyes-closed EEG data were recorded from 60 depressed patients. A phase synchronization measure was used to estimate functional connectivity between all pairs of the EEG channels in the delta (1-4 Hz), theta (4-8 Hz), alpha (8-13 Hz), and beta (13-30 Hz) frequency bands. To quantify the local value of functional connectivity, 2 graph theory metrics, degree, and clustering coefficient (CC), were measured. Moreover, Lempel-Ziv complexity (LZC) and fuzzy entropy (FuzzyEn) were used to measure the complexity of the EEG signal.

RESULTS

Through correlation analysis, a significant negative relationship was found between graph metrics and depression severity in the alpha band. This association was strongly positive for the complexity measures in alpha and delta bands. Also, the linear regression model represented a substantial performance of depression severity prediction based on EEG features of the alpha band (r = 0.839; P < .0001, root mean square error score of 7.69).

CONCLUSION

We found that the brain activity of patients with depression was related to depression severity. Abnormal brain activity reflects an increase in the severity of depression. The presented regression model provides a quantitative depression severity prediction, which can inform the development of EEG state and exhibit potential desirable application for the medical treatment of the depressive disorder.

Automated Detection of Sleep Stages Using Deep Learning Techniques: A Systematic Review of the Last Decade (2010–2020)

Sleep is vital for one’s general well-being, but it is often neglected, which has led to an increase in sleep disorders worldwide. Indicators of sleep disorders, such as sleep interruptions, extreme daytime drowsiness, or snoring, can be detected with sleep analysis. However, sleep analysis relies on visuals conducted by experts, and is susceptible to inter- and intra-observer variabilities. One way to overcome these limitations is to support experts with a programmed diagnostic tool (PDT) based on artificial intelligence for timely detection of sleep disturbances. Artificial intelligence technology, such as deep learning (DL), ensures that data are fully utilized with low to no information loss during training. This paper provides a comprehensive review of 36 studies, published between March 2013 and August 2020, which employed DL models to analyze overnight polysomnogram (PSG) recordings for the classification of sleep stages. Our analysis shows that more than half of the studies employed convolutional neural networks (CNNs) on electroencephalography (EEG) recordings for sleep stage classification and achieved high performance. Our study also underscores that CNN models, particularly one-dimensional CNN models, are advantageous in yielding higher accuracies for classification. More importantly, we noticed that EEG alone is not sufficient to achieve robust classification results. Future automated detection systems should consider other PSG recordings, such as electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG) signals, along with input from human experts, to achieve the required sleep stage classification robustness. Hence, for DL methods to be fully realized as a practical PDT for sleep stage scoring in clinical applications, inclusion of other PSG recordings, besides EEG recordings, is necessary. In this respect, our report includes methods published in the last decade, underscoring the use of DL models with other PSG recordings, for scoring of sleep stages.

A Novel Method for Sleep-Stage Classification Based on Sonification of Sleep Electroencephalogram Signals Using Wavelet Transform and Recurrent Neural Network

INTRODUCTION

Visual sleep-stage scoring is a time-consuming technique that cannot extract the nonlinear characteristics of electroencephalogram (EEG). This article presents a novel method for sleep-stage differentiation based on sonification of sleep-EEG signals using wavelet transform and recurrent neural network (RNN).

METHODS

Two RNNs were designed and trained separately based on a database of classical guitar pieces and Kurdish tanbur Makams using a long short-term memory model. Moreover, discrete wavelet transform and wavelet packet decomposition were used to determine the association between the EEG signals and musical pitches. Continuous wavelet transform was applied to extract musical beat-based features from the EEG. Then, the pretrained RNN was used to generate music. To test the proposed model, 11 sleep EEGs were mapped onto the guitar and tanbur frequency intervals and presented to the pretrained RNN. Next, the generated music was randomly presented to 2 neurologists.

RESULTS

The proposed model classified the sleep stages with an accuracy of >81% for tanbur and more than 93% for guitar musical pieces. The inter-rater reliability measured by Cohen's kappa coefficient (κ) revealed good reliability for both tanbur (κ = 0.64, p < 0.001) and guitar musical pieces (κ = 0.85, p < 0.001).

CONCLUSION

The present EEG sonification method leads to valid sleep staging by clinicians. The method could be used on various EEG databases for classification, differentiation, diagnosis, and treatment purposes. Real-time EEG sonification can be used as a feedback tool for replanning of neurophysiological functions for the management of many neurological and psychiatric disorders in the future.

Review on Biomedical Sensors, Technologies and Algorithms for Diagnosis of Sleep Disordered Breathing: Comprehensive Survey

This paper provides a comprehensive review of available technologies for measurements of vital physiology related parameters that cause sleep disordered breathing (SDB). SDB is a chronic disease that may lead to several health problems and increase the risk of high blood pressure and even heart attack. Therefore, the diagnosis of SDB at an early stage is very important. The essential primary step before diagnosis is measurement. Vital health parameters related to SBD might be measured through invasive or non-invasive methods. Nowadays, with respect to increase in aging population, improvement in home health management systems is needed more than even a decade ago. Moreover, traditional health parameter measurement techniques such as polysomnography are not comfortable and introduce additional costs to the consumers. Therefore, in modern advanced self-health management devices, electronics and communication science are combined to provide appliances that can be used for SDB diagnosis, by monitoring a patient's physiological parameters with more comfort and accuracy. Additionally, development in machine learning algorithms provides accurate methods of analysing measured signals. This paper provides a comprehensive review of measurement approaches, data transmission, and communication networks, alongside machine learning algorithms for sleep stage classification, to diagnose SDB.

EEG-based deep learning model for the automatic detection of clinical depression

Clinical depression is a neurological disorder that can be identified by analyzing the Electroencephalography (EEG) signals. However, the major drawback in using EEG to accurately identify depression is the complexity and variation that exist in the EEG of a depressed individual. There are several strategies for automated depression diagnosis, but they all have flaws, which make the diagnostic task inaccurate. In this paper, a deep model is designed in which an integration of Convolution Neural Network (CNN) and Long Short Term Memory (LSTM) is implemented for the detection of depression. CNN and LSTM are used to learn the local characteristics and the EEG signal sequence, respectively. In the deep learning model, filters in the convolution layer are convolved with the input signal to generate feature maps. All the extracted features are given to the LSTM for it to learn the different patterns in the signal, after which the classification is performed using fully connected layers. LSTM has memory cells to remember the essential features for a long time. It also has different functions to update the weights during training. Testing of the model was done by random splitting technique and obtained 99.07% and 98.84% accuracies for the right and left hemispheres EEG signals, respectively.

An Emotion Assessment of Stroke Patients by Using Bispectrum Features of EEG Signals

Emotion assessment in stroke patients gives meaningful information to physiotherapists to identify the appropriate method for treatment. This study was aimed to classify the emotions of stroke patients by applying bispectrum features in electroencephalogram (EEG) signals. EEG signals from three groups of subjects, namely stroke patients with left brain damage (LBD), right brain damage (RBD), and normal control (NC), were analyzed for six different emotional states. The estimated bispectrum mapped in the contour plots show the different appearance of nonlinearity in the EEG signals for different emotional states. Bispectrum features were extracted from the alpha (8-13) Hz, beta (13-30) Hz and gamma (30-49) Hz bands, respectively. The k-nearest neighbor (KNN) and probabilistic neural network (PNN) classifiers were used to classify the six emotions in LBD, RBD and NC. The bispectrum features showed statistical significance for all three groups. The beta frequency band was the best performing EEG frequency-sub band for emotion classification. The combination of alpha to gamma bands provides the highest classification accuracy in both KNN and PNN classifiers. Sadness emotion records the highest classification, which was 65.37% in LBD, 71.48% in RBD and 75.56% in NC groups.

Motor execution reduces EEG signals complexity: Recurrence quantification analysis study

The development of new approaches to detect motor-related brain activity is key in many aspects of science, especially in brain-computer interface applications. Even though some well-known features of motor-related electroencephalograms have been revealed using traditionally applied methods, they still lack a robust classification of motor-related patterns. Here, we introduce new features of motor-related brain activity and uncover hidden mechanisms of the underlying neuronal dynamics by considering event-related desynchronization (ERD) of μ-rhythm in the sensorimotor cortex, i.e., tracking the decrease of the power spectral density in the corresponding frequency band. We hypothesize that motor-related ERD is associated with the suppression of random fluctuations of μ-band neuronal activity. This is due to the lowering of the number of active neuronal populations involved in the corresponding oscillation mode. In this case, we expect more regular dynamics and a decrease in complexity of the EEG signal recorded over the sensorimotor cortex. In order to support this, we apply measures of signal complexity by means of recurrence quantification analysis (RQA). In particular, we demonstrate that certain RQA quantifiers are very useful to detect the moment of movement onset and, therefore, are able to classify the laterality of executed movements.

A review of automated sleep stage scoring based on physiological signals for the new millennia

BACKGROUND AND OBJECTIVE

Sleep is an important part of our life. That importance is highlighted by the multitude of health problems which result from sleep disorders. Detecting these sleep disorders requires an accurate interpretation of physiological signals. Prerequisite for this interpretation is an understanding of the way in which sleep stage changes manifest themselves in the signal waveform. With that understanding it is possible to build automated sleep stage scoring systems. Apart from their practical relevance for automating sleep disorder diagnosis, these systems provide a good indication of the amount of sleep stage related information communicated by a specific physiological signal.

METHODS

This article provides a comprehensive review of automated sleep stage scoring systems, which were created since the year 2000. The systems were developed for Electrocardiogram (ECG), Electroencephalogram (EEG), Electrooculogram (EOG), and a combination of signals.

RESULTS

Our review shows that all of these signals contain information for sleep stage scoring.

CONCLUSIONS

The result is important, because it allows us to shift our research focus away from information extraction methods to systemic improvements, such as patient comfort, redundancy, safety and cost.

Introducing chaotic codes for the modulation of code modulated visual evoked potentials (c-VEP) in normal adults for visual fatigue reduction

Code modulated Visual Evoked Potentials (c-VEP) based BCI studies usually employ m-sequences as a modulating codes for their broadband spectrum and correlation property. However, subjective fatigue of the presented codes has been a problem. In this study, we introduce chaotic codes containing broadband spectrum and similar correlation property. We examined whether the introduced chaotic codes could be decoded from EEG signals and also compared the subjective fatigue level with m-sequence codes in normal subjects. We generated chaotic code from one-dimensional logistic map and used it with conventional 31-bit m-sequence code. In a c-VEP based study in normal subjects (n = 44, 21 females) we presented these codes visually and recorded EEG signals from the corresponding codes for their four lagged versions. Canonical correlation analysis (CCA) and spatiotemporal beamforming (STB) methods were used for target identification and comparison of responses. Additionally, we compared the subjective self-declared fatigue using VAS caused by presented m-sequence and chaotic codes. The introduced chaotic code was decoded from EEG responses with CCA and STB methods. The maximum total accuracy values of 93.6 ± 11.9% and 94 ± 14.4% were achieved with STB method for chaotic and m-sequence codes for all subjects respectively. The achieved accuracies in all subjects were not significantly different in m-sequence and chaotic codes. There was significant reduction in subjective fatigue caused by chaotic codes compared to the m-sequence codes. Both m-sequence and chaotic codes were similar in their accuracies as evaluated by CCA and STB methods. The chaotic codes significantly reduced subjective fatigue compared to the m-sequence codes.

A Deep Learning Model for Automated Sleep Stages Classification Using PSG Signals

Sleep disorder is a symptom of many neurological diseases that may significantly affect the quality of daily life. Traditional methods are time-consuming and involve the manual scoring of polysomnogram (PSG) signals obtained in a laboratory environment. However, the automated monitoring of sleep stages can help detect neurological disorders accurately as well. In this study, a flexible deep learning model is proposed using raw PSG signals. A one-dimensional convolutional neural network (1D-CNN) is developed using electroencephalogram (EEG) and electrooculogram (EOG) signals for the classification of sleep stages. The performance of the system is evaluated using two public databases (sleep-edf and sleep-edfx). The developed model yielded the highest accuracies of 98.06%, 94.64%, 92.36%, 91.22%, and 91.00% for two to six sleep classes, respectively, using the sleep-edf database. Further, the proposed model obtained the highest accuracies of 97.62%, 94.34%, 92.33%, 90.98%, and 89.54%, respectively for the same two to six sleep classes using the sleep-edfx dataset. The developed deep learning model is ready for clinical usage, and can be tested with big PSG data.

Using Wearable ECG/PPG Sensors for Driver Drowsiness Detection Based on Distinguishable Pattern of Recurrence Plots

This paper aims to investigate the robust and distinguishable pattern of heart rate variability (HRV) signals, acquired from wearable electrocardiogram (ECG) or photoplethysmogram (PPG) sensors, for driver drowsiness detection. As wearable sensors are so vulnerable to slight movement, they often produce more noise in signals. Thus, from noisy HRV signals, we need to find good traits that differentiate well between drowsy and awake states. To this end, we explored three types of recurrence plots (RPs) generated from the R–R intervals (RRIs) of heartbeats: Bin-RP, Cont-RP, and ReLU-RP. Here Bin-RP is a binary recurrence plot, Cont-RP is a continuous recurrence plot, and ReLU-RP is a thresholded recurrence plot obtained by filtering Cont-RP with a modified rectified linear unit (ReLU) function. By utilizing each of these RPs as input features to a convolutional neural network (CNN), we examined their usefulness for drowsy/awake classification. For experiments, we collected RRIs at drowsy and awake conditions with an ECG sensor of the Polar H7 strap and a PPG sensor of the Microsoft (MS) band 2 in a virtual driving environment. The results showed that ReLU-RP is the most distinct and reliable pattern for drowsiness detection, regardless of sensor types (i.e., ECG or PPG). In particular, the ReLU-RP based CNN models showed their superiority to other conventional models, providing approximately 6–17% better accuracy for ECG and 4–14% for PPG in drowsy/awake classification.

A Novel Multi-Class EEG-Based Sleep Stage Classification System

Sleep stage classification is one of the most critical steps in effective diagnosis and the treatment of sleep-related disorders. Visual inspection undertaken by sleep experts is a time-consuming and burdensome task. A computer-assisted sleep stage classification system is thus essential for both sleep-related disorders diagnosis and sleep monitoring. In this paper, we propose a system to classify the wake and sleep stages with high rates of sensitivity and specificity. The EEG signals of 25 subjects with suspected sleep-disordered breathing, and the EEG signals of 20 healthy subjects from three data sets are used. Every EEG epoch is decomposed into eight subband epochs each of which has a frequency band pertaining to one EEG rhythm (i.e., delta, theta, alpha, sigma, beta 1, beta 2, gamma 1, or gamma 2). Thirteen features are extracted from each subband epoch. Therefore, 104 features are totally obtained for every EEG epoch. The Kruskal-Wallis test is used to examine the significance of the features. Non-significant features are discarded. The minimal-redundancy-maximal-relevance feature selection algorithm is then used to eliminate redundant and irrelevant features. The features selected are classified by a random forest classifier. To set the system parameters and to evaluate the system performance, nested 5-fold cross-validation and subject cross-validation are performed. The performance of our proposed system is evaluated for different multi-class classification problems. The minimum overall accuracy rates obtained are 95.31% and 86.64% for nested 5-fold and subject cross-validation, respectively. The system performance is promising in terms of the accuracy, sensitivity, and specificity rates compared with the ones of the state-of-the-art systems. The proposed system can be used in health care applications with the aim of improving sleep stage classification.

Multiscale permutation Rényi entropy and its application for EEG signals

There is considerable interest in analyzing the complexity of electroencephalography (EEG) signals. However, some traditional complexity measure algorithms only quantify the complexities of signals, but cannot discriminate different signals very well. To analyze the complexity of epileptic EEG signals better, a new multiscale permutation Rényi entropy (MPEr) algorithm is proposed. In this algorithm, the coarse-grained procedure is introduced by using weighting-averaging method, and the weighted factors are determined by analyzing nonlinear signals. We apply the new algorithm to analyze epileptic EEG signals. The experimental results show that MPEr algorithm has good performance for discriminating different EEG signals. Compared with permutation Rényi entropy (PEr) and multiscale permutation entropy (MPE), MPEr distinguishes different EEG signals successfully. The proposed MPEr algorithm is effective and has good applications prospects in EEG signals analysis.

EEG-Based Automatic Sleep Staging Using Ontology and Weighting Feature Analysis

Sleep staging is considered as an effective indicator for auxiliary diagnosis of sleep diseases and related psychiatric diseases, so it attracts a lot of attention from sleep researchers. Nevertheless, sleep staging based on visual inspection of tradition is subjective, time-consuming, and error-prone due to the large bulk of data which have to be processed. Therefore, automatic sleep staging is essential in order to solve these problems. In this article, an electroencephalogram- (EEG-) based scheme that is able to automatically classify sleep stages is proposed. Firstly, EEG data are preprocessed to remove artifacts, extract features, and normalization. Secondly, the normalized features and other context information are stored using an ontology-based model (OBM). Thirdly, an improved method of self-adaptive correlation analysis is designed to select the most effective EEG features. Based on these EEG features and weighting features analysis, the improved random forest (RF) is considered as the classifier to achieve the classification of sleep stages. To investigate the classification ability of the proposed method, several sets of experiments are designed and conducted to classify the sleep stages into two, three, four, and five states. The accuracy of five-state classification is 89.37%, which is improved compared to the accuracy using unimproved RF (84.37%) or previously reported classifiers. In addition, a set of controlled experiments is executed to verify the effect of the number of sleep segments (epochs) on the classification, and the results demonstrate that the proposed scheme is less affected by the sleep segments.

Integration of 24 Feature Types to Accurately Detect and Predict Seizures Using Scalp EEG Signals

The neurological disorder epilepsy causes substantial problems to the patients with uncontrolled seizures or even sudden deaths. Accurate detection and prediction of epileptic seizures will significantly improve the life quality of epileptic patients. Various feature extraction algorithms were proposed to describe the EEG signals in frequency or time domains. Both invasive intracranial and non-invasive scalp EEG signals have been screened for the epileptic seizure patterns. This study extracted a comprehensive list of 24 feature types from the scalp EEG signals and found 170 out of the 2794 features for an accurate classification of epileptic seizures. An accuracy (Acc) of 99.40% was optimized for detecting epileptic seizures from the scalp EEG signals. A balanced accuracy (bAcc) was calculated as the average of sensitivity and specificity and our seizure detection model achieved 99.61% in bAcc. The same experimental procedure was applied to predict epileptic seizures in advance, and the model achieved Acc = 99.17% for predicting epileptic seizures 10 s before happening.

Nonlinear Recurrent Dynamics and Long-Term Nonstationarities in EEG Alpha Cortical Activity: Implications for Choosing Adequate Segment Length in Nonlinear EEG Analyses

Nonlinear analysis of EEG recordings allows detection of characteristics that would probably be neglected by linear methods. This study aimed to determine a suitable epoch length for nonlinear analysis of EEG data based on its recurrence rate in EEG alpha activity (electrodes Fz, Oz, and Pz) from 28 healthy and 64 major depressive disorder subjects. Two nonlinear metrics, Lempel-Ziv complexity and scaling index, were applied in sliding windows of 20 seconds shifted every 1 second and in nonoverlapping windows of 1 minute. In addition, linear spectral analysis was carried out for comparison with the nonlinear results. The analysis with sliding windows showed that the cortical dynamics underlying alpha activity had a recurrence period of around 40 seconds in both groups. In the analysis with nonoverlapping windows, long-term nonstationarities entailed changes over time in the nonlinear dynamics that became significantly different between epochs across time, which was not detected with the linear spectral analysis. Findings suggest that epoch lengths shorter than 40 seconds neglect information in EEG nonlinear studies. In turn, linear analysis did not detect characteristics from long-term nonstationarities in EEG alpha waves of control subjects and patients with major depressive disorder patients. We recommend that application of nonlinear metrics in EEG time series, particularly of alpha activity, should be carried out with epochs around 60 seconds. In addition, this study aimed to demonstrate that long-term nonlinearities are inherent to the cortical brain dynamics regardless of the presence or absence of a mental disorder.

ALCOHOLIC INDEX USING NON-LINEAR FEATURES EXTRACTED FROM DIFFERENT FREQUENCY BANDS

Alcoholism is a complex condition that mainly disturbs the neuronal networks in Central Nervous System (CNS). This disorder not only disturbs the brain, but also affects the behavior, emotions, and cognitive judgements. Electroencephalography (EEG) is a valuable tool to examine the neuropsychiatric disorders like alcoholism. The EEG is a well-established modality to diagnose the electrical activity produced by the populations of neurons in cerebral cortex. However, EEG signals are non-linear in nature; hence very challenging to interpret the valuable information from them using linear methods. Thus, using non-linear methods to analyze EEG signals can be beneficial in order to predict the brain signals condition. This paper presents a computer-aided diagnostic method for the detection of alcoholic EEG signals from normal by employing the non-linear techniques. First, the EEG signals are subjected to six levels of Wavelet Packet Decomposition (WPD) to obtain seven wavebands (delta ([Formula: see text]), theta ([Formula: see text]), lower alpha (la), upper alpha (ua), lower beta (lb), upper beta (ub), lower gamma (lg)). From each wavebands (activity bands), 19 non-linear features such as Recurrence Quantification Analysis (RQA) ([Formula: see text]), Approximate Entropy ([Formula: see text]), Energy ([Formula: see text]), Fractal Dimension (FD) ([Formula: see text]), Permutation Entropy ([Formula: see text]), Detrended Fluctuation Analysis ([Formula: see text]), Hurst Exponent ([Formula: see text]), Largest Lyapunov Exponent ([Formula: see text]), Sample Entropy ([Formula: see text]), Shannon’s Entropy ([Formula: see text]), Renyi’s entropy ([Formula: see text]), Tsalli’s entropy ([Formula: see text]), Fuzzy entropy ([Formula: see text]), Wavelet entropy ([Formula: see text]), Kolmogorov–Sinai entropy ([Formula: see text]), Modified Multiscale Entropy ([Formula: see text]), Hjorth’s parameters (activity ([Formula: see text]), mobility ([Formula: see text]), and complexity ([Formula: see text])) are extracted. The extracted features are then ranked using Bhattacharyya, Entropy, Fuzzy entropy-based Max-Relevancy and Min-Redundancy (mRMR), Receiver Operating Characteristic (ROC), [Formula: see text]-test, and Wilcoxon. These ranked features are given to train Support Vector Machine (SVM) classifier. The SVM classifier with radial basis function (RBF) achieved 95.41% accuracy, 93.33% sensitivity and 97.50% specificity using four non-linear features ranked by Wilcoxon method. In addition, an integrated index called Alcoholic Index (ALCOHOLI) is developed using highly ranked two features for identification of normal and alcoholic EEG signals using a single number. This system is rapid, efficient, and inexpensive and can be employed as an EEG analysis assisting system by clinicians in the detection of alcoholism. In addition, the proposed system can be used in rehabilitation centers to evaluate person with alcoholism over time and observe the outcome of treatment provided for reducing or reversing the impact of the condition on the brain.

Automatic sleep staging using ear-EEG

BACKGROUND

Sleep and sleep quality assessment by means of sleep stage analysis is important for both scientific and clinical applications. Unfortunately, the presently preferred method, polysomnography (PSG), requires considerable expert assistance and significantly affects the sleep of the person under observation. A reliable, accurate and mobile alternative to the PSG would make sleep information much more readily available in a wide range of medical circumstances.

NEW METHOD

Using an already proven method, ear-EEG, in which electrodes are placed inside the concha and ear canal, we measure cerebral activity and automatically score the sleep into up to five stages. These results are compared to manual scoring by trained clinicians, based on a simultaneously recorded PSG.

RESULTS

The correspondence between manually scored sleep, based on the PSG, and the automatic labelling, based on ear-EEG data, was evaluated using Cohen's kappa coefficient. Kappa values are in the range 0.5-0.8, making ear-EEG relevant for both scientific and clinical applications. Furthermore, a sleep-wake classifier with leave-one-out cross validation yielded specificity of 0.94 and sensitivity of 0.52 for the sleep stage.

COMPARISON WITH EXISTING METHOD(S)

Ear-EEG based scoring has clear advantages when compared to both the PSG and other mobile solutions, such as actigraphs. It is far more mobile, and potentially cheaper than the PSG, and the information on sleep stages is far superior to a wrist-based actigraph, or other devices based solely on body movement.

CONCLUSIONS

This study shows that ear-EEG recordings carry information about sleep stages, and indicates that automatic sleep staging based on ear-EEG can classify sleep stages with a level of accuracy that makes it relevant for both scientific and clinical sleep assessment.