Contactless OSAHS Respiration and Sleep Stage Classification Using Monitored Snoring Events With Heterogeneous Multiscale Selective Distillation for Internet of Medical Things

Automatic Sleep Stage Classification using Weighted ELM and PSO on Imbalanced Data from Single Lead ECG

The goal of this work was to describe an automated system for classification of human sleep stages. Classification of sleep stages is an important problem of diagnosis and treatment of human sleep disorders. The developed classification method is based on calculation of characteristics of the main sleep rhythms. It uses hidden Markov models. The method is highly accurate and provides reliable identification of the main stages of sleep. The results of automatic classification are in good agreement with the results of sleep stage identification performed by an expert somnologist using Rechtschaffen and Kales rules. This substantiates the applicability of the developed classification system to clinical diagnosis.

Background: Sleep disorders become one of the early warnings of Non-Communicable Diseases (NCDs).In sleep stage classification process one of the important stages to be sleep score recording. Generally the first step of diagnosis of sleep disorder is to be Polysomnography (PSG) test.The PSG test is a formal method to diagnose sleep disorders, during this test we have considered many biomedical signals such as electroencephalogram (EEG),electrooculogram (EOG) and electromyogram (EMG).Sleep Stage classification (SSC) process is time taking and there must be a presence of sleep experts or technicians definitely to be stay with subject through the whole recording time period, which is somehow overburden for clinicians and it may hamper sometime to record the correct results from subjects. For that reason now researchers obtained Automatic Sleep Stage Classification (ASSC) methods in order to find disturbances during sleep and it's quite faster and efficient in towards data recording accuracy from PSG signals.Method: In this study, we have proposed an alternative approach for sleep stage scoring by considering different age subjects from same gender with their optimal features. In this proposed study we have considered dual channels of EEG signals such as C4-A1 and O2-A1. In data pre-processing stage, the datasets were analyzed and normalized using feature extraction and feature selection methods. The main important part of this research work is to be comparing between dual channels and its accuracy of best discrimination in between wake and sleep stages. In addition, we have obtained three base classifier such as support vector machines (SVM), decision tree (DT) and K-nearest neighbors (KNN). In addition we have also adopted ensemble classifier (Boosting), to make a proper comparison the classification performances in between them. For validation purpose in between training data and test data, we have used 10-fold cross validation techniques.Results: In this study, we have made a comparison the performances in terms channel effectiveness and classification algorithm effectiveness with regard to discriminate in between the sleep stages. As per outcome from the proposed system the SVM classification techniques achieved best accuracy in comparable to other classifiers.Regading to channel effectiveness, C4-A1 recording is more appropriate for sleep stage scoring.Discussion and Conclusion: As per the related research work in this field , the introduced approach in the present study achieved an acceptable performance in sleep scoring in order to classifying wake stage and sleep stage from dual channels of EEG signals. Our experiment design compares the accuracy of classification in between two channels and find out which channel recordings and classification techniques to be most effective in towards classifying sleep stages. In future its performance to be increase through proper enhancing through different intelligent techniques in related to process of diagnosing and treatment of sleep disorders.

Joint ensemble empirical mode decomposition and tunable Q factor wavelet transform based sleep stage classifications

Backgrounds and Objectives: Sleep-related disorders are critical diseases and they need to be proper diagnosed as early as possible. The major difficulty is that fewer medical experiments are available in remote locations to diagnose different types of sleep disorders. This research work presents an automated classification of sleep stages using deep learning techniques to diagnosis multiple sleep diseases from single-channel and different combinations of channels of EEG signals. Methods: In this proposed work, we involve three different forms of signal inputs for the automatic classification of sleep stages from EEG signals. The proposed research work presents a one-dimensional convolution neural network (1D-CNN) model for multiple sleep stages because of its high robustness for automatically classifying the sleep stages from brain signals without involving any types of feature extraction/selection, which is one of the challenging processes in the earlier literature. The proposed model contained seven convolution layers followed by two fully connected layers. The main objective for designing such a custom deep neural network is to improve classification accuracy performance with less number of learnable parameters. Results: The proposed model has used two subgroups of the ISRUC-Sleep dataset. We also obtain a k-cross-fold validation approach over the subjects, which ensure that there is no possible contamination in between training and testing. The experimental results for this proposed model for classification of five classes of sleep stages (wake, non-rapid eye movement N1-N3, and rapid eye movement). The proposed model was evaluated by classification accuracy, precision, sensitivity, F1score, and Cohen's Kappa score. The proposed 1D-CNN architecture achieved the highest classification accuracy of 95.85%(C3-A2),94.11%(O1-A2), and 97.22%(C3-A2+O1-A2) using the SG-I dataset, similarly, the same model reported accuracy using the SG-II dataset of 95.73%(C3-A2),94.02%(O1-A2), and 95.06% (C3-A2+O1-A2). Conclusions: Our proposed methodology is efficient and effective for multiple sleep staging. The proposed 1D-CNN model is ready for clinical purposes and can be managed with a huge number of polysomnography data.

Sleep is closely related to physical and mental health and quality of life, and accurately evaluating sleep quality remains a major research topic in related fields. Conventional methods of sleep quality evaluation involve the use of polysomnography (PSG), which continuously records physiological changes during sleep. With the recorded data, sleep quality, and sleep stage or sleep-related disorders can be diagnosed via manual inspection by trained experts. However, the practice is time-consuming, labor-intensive and yields high inter-and intra-rater variability. To overcome such disadvantages, machine learning or deep learning-based automated techniques have been recently proposed. Most of the existing methods, however, can only classify sleep stage or the presence of apnea, not both. This study proposes a novel method that allows simultaneous classification of both sleep stage and apnea, through multioutput classification using a single deep learning model to comprehensively evaluate sleep quality. PSG recordings from a total of 98 subjects from the ISRUC-Sleep dataset subgroup 1 were analyzed in this study, and the proposed model was trained to classify sleep stage and apnea using three EEG channels. The proposed model showed an accuracy and a micro average f1 score of 0.99 for the classification of apnea, and an accuracy and a micro average f1 score of 0.69 for the classification of the sleep stages. The proposed model provides a new paradigm in sleep deep learning research by enabling comprehensive evaluation of sleep quality, hence could help sleep experts to make clinical decisions.

Maintaining human physical and mental health depends on sleep; insufficient sleep results in illness. Many deep learning (DL) and machine learning (ML) based sleep stage classification (SSC) algorithms have been proposed in the last ten years. However, insufficient feature uniqueness, intricate deep learning architectures, a great deal of hyperparameter adjustment, and low accuracy in classifying sleep stages make SSC difficult. This paper presents SSC based on spectral, texture, and temporal features of electroencephalogram (EEG) features and a lightweight Deep Convolution Neural Network (DCNN) and Long Short-Term Memory (LSTM). The DCNN helps to improve the correlation and connectivity features of EEG, and the LSTM helps to boost the temporal depiction and long-term connectivity of EEG features. It uses Wavelet Packet Transform (WPT) based soft thresholding to minimize noise and artifacts in the EEG signal. The improved Spider Monkey Optimization (ISMO) algorithm selects the decisive features from the multiple EEG features. The suggested WPT-ISMO-DCNN-LSTM-based SSC scheme's effectiveness is estimated on the sleep-European Data Format (Sleep-EDF) dataset based on accuracy, recall, precision, F1-score, trainable parameters, and recognition time. The WPT-ISMO-DCNN-LSTM-based SSC scheme provides an accuracy of 99%, precision of 1, Kohen's Kappa rate of 0.9878, 0.9921, recall of 0.99, and F1-score of 0.99, outperforming the existing state of the art. The WPT-ISMO-DCNN-LSTM provides an accuracy of 99% for 2-class SSC, 99.70% for 3-class SSC, 98.50% for 4-class SSC, 99.30% for 5-class SSC, and 99% for 6-class SSC for 350 features. The proposed algorithm offers 5.9 M trainable parameters and a total training time of 498 s for a 6-class SSC.

The most important part of sleep quality assessment is the classification of sleep stages, which helps to diagnose sleep-related disease. In the traditional sleep staging method, subjects have to spend a night in the sleep clinic for recording polysomnogram. Sleep expert classifies the sleep stages by monitoring the signals, which is time consuming and frustrating task and can be affected by human error. New studies propose fully automated techniques for classifying sleep stages that makes sleep scoring possible at home. Despite comprehensive studies have been presented in this field the classification results have not yet reached the gold standard due to the concentration on the use of a limited source of information such as single channel EEG. Therefore, this article introduces a new method for fusing two sources of information, including electroencephalogram (EEG) and electrooculogram (EOG), to achieve promising results in the classification of sleep stages. In the proposed method, extracted features from the EEG and EOG signals, are divided into two feature sets consisting of the EEG features and fused features of EEG and EOG. Then, each feature set transformed into a horizontal visibility graph (HVG). The images of the HVG are produced in a novel framework and classified by proposed transfer learning convolutional neural network for data fusion (TLCNN-DF). Employing transfer learning at the training stage of the model has accelerated the training process of the CNN and improved the performance of the model. The proposed algorithm is used to classify the Sleep-EDF and Sleep-EDFx benchmark datasets. The algorithm can classify the Sleep-EDF dataset with an accuracy of 93.58% and Cohen’s kappa coefficient of 0.899. The results show proposed method can achieve superior performance compared to state-of-the-art studies on classification of sleep stages. Furthermore, it can attain reliable results as an alternative to conventional sleep staging.

Sleep stage classification is crucial for sleep state monitoring and health interventions. In accordance with the standards prescribed by the American Academy of Sleep Medicine, a sleep episode follows a specific structure comprising five distinctive sleep stages that collectively form a sleep cycle. Typically, this cycle repeats about five times, providing an insightful portrayal of the subject’s physiological attributes. The progress of deep learning and advanced domain generalization methods allows automatic and even adaptive sleep stage classification. However, applying models trained with visible subject data to invisible subject data remains challenging due to significant individual differences among subjects. Motivated by the periodic category-complete structure of sleep stage classification, we propose a Structure Incentive Domain Adversarial learning (SIDA) method that combines the sleep stage classification method with domain generalization to enable cross-subject sleep stage classification. SIDA includes individual domain discriminators for each sleep stage category to decouple subject dependence differences among different categories and fine-grained learning of domain-invariant features. Furthermore, SIDA directly connects the label classifier and domain discriminators to promote the training process. Experiments on three benchmark sleep stage classification datasets demonstrate that the proposed SIDA method outperforms other state-of-the-art sleep stage classification and domain generalization methods and achieves the best cross-subject sleep stage classification results.

In this paper, we analyze the performance of electrocardiogram (ECG) signal compression by comparing original and reconstructed signal on two problems. First, automatic sleep stage classification based on ECG signal; second, arrhythmia classification. An effective ECG signal compression method based on two-dimensional wavelet transform which employs set partitioning in hierarchical trees (SPIHT) and beat reordering technique used to compress the ECG signal. This method utilizes the redundancy between adjacent samples and adjacent beats. Beat reordering rearranges beat order in 2D (2 dimension) ECG array based on the similarity between adjacent beats. The experimental results show that the proposed method yields relatively low distortion at high compression rate. The experimental results also show that the accuracy of sleep stage classification and arrhythmia classification using reconstructed ECG signal from proposed method is comparable to the original signal. The proposed method preserved signal characteristics for the automatic sleep stage and arrhythmia classification problems.

The present work mainly focused on automatic classification of the sleep stages classification from the different medical-conditioned subjects under American Academy of Sleep Medicine (AASM). The present research study was used single individual channel for classifying two-state sleep stage classification problem in between wake versus Sleep. During first part, we have conducted experiments such as acquisition of data from participated subjects, preprocessing the raw signal to remove the irrelevant artefacts and muscle movements from recorded sleep data, extract quantitative features obtained from EEG signal and also used feature selection techniques to choose the suitable features which is most useful for characterizing the abnormality pattern. In second step, we have used two machine learning classifiers such as support vector machine (SVM) and K-nearest neighbor (KNN). The proposed scheme was conducted through six different medical conditions subjects with considering both affected sleep diseases subjects and healthy subjects and their one session and two session recordings. The obtained results demonstrated that the proposed methodologies support to sleep experts for accurately measure the irregularities occurred during sleep and also helps the clinicians to evaluate the presence and criticality of sleep related disorders.

Sleep disorders have been increasingly prevalent, and the necessity for sleep stage classification, a pivotal component in the diagnosis of sleep disorders, is also rising. The majority of sleep stage classifications employ multichannel biosignals derived from polysomnography. However, this approach is impractical, prompting the exploration of sleep stage models that utilize a single channel. In this study, we propose a model that automatically classifies sleep stages using a single ECG channel that is user-friendly and easily wearable. The proposed method integrates convolutional neural networks and transformer structures to learn both local and global information for sleep stage classification. In the context of 4-stage sleep stage classification (i.e., wake, light sleep, deep sleep, and rapid eye movement), the method attained accuracies of 76.12% and 63.42% on ISRUC-1 and SHHS-1, respectively, thereby demonstrating superior performance in comparison to baseline models. The proposed framework may offer significant potential for automatic sleep stage classification and may aid in the accurate diagnosis of sleep disorders.Clinical relevance- This is straightforward and can be utilized to diagnose sleep disorders, such as sleep apnea, by enhancing the precision of sleep stage classification.

Sleep stage classification is crucial for sleep quality assessment and disease diagnosis. Although some recent studies have made great strides in sleep stage classification performance, direct application to multi-modal sleep data with cross-domain distributional variations still poses challenges: 1) How to retain the sleep knowledge acquired by the model from the source domain during cross-domain adaptation to avoid catastrophic forgetting. 2) How to evaluate the contribution of different modalities in identifying specific sleep stages to serve test-time adaptation (TTA). 3) How to dynamically adapt the sleep model to different distribution shift in data domains of different subjects. To address these challenges, we propose an Adaptive Test-Time Adaptation (ATTA) method, a multi-modal test-time adaptation method for sleep stage classification. Specifically, the intra-modal retained-adaptive module is proposed for adapting to the target domain data while retaining the sleep knowledge acquired from the source domain to avoid catastrophic forgetting. The inter-modal contribution assessment module is designed to adaptively assess the contribution of each modality to the identification of specific sleep stages. Furthermore, the adaptive learning rate strategy utilizes a memory bank to record data from different subjects during testing, and based on this, it measures the differences between the target subject and those in the memory bank. According to the difference, the model adapts to the subject samples with different learning rates. We conduct experiments on mutual migration on two sleep datasets, SleepEDF and SHHS. The results show that our ATTA method outperforms state-of-the-art baselines in sleep stage classification.

Increasingly prevalent sleep disorders worldwide significantly affect the well-being of individuals. Sleep disorder can be detected by dividing sleep into different stages. Hence, the accurate classification of sleep stages is crucial for detecting sleep disorders. The use of machine learning techniques on physiological signals has shown promising results in the automatic classification of sleep stages. The integration of information from multichannel physiological signals has shown to further enhance the accuracy of such classification. Existing literature reviews focus on studies utilising a single channel of EEG signals for sleep stage classification. However, other review studies focus on models developed for sleep stage classification, utilising either a single channel of physiological signals or a combination of various physiological signals. This review focuses on the classification of sleep stages through the integration of combined multichannel physiological signals and machine learning methods. We conducted a comprehensive review spanning from the year 2000 to 2023, aiming to provide a thorough and up-to-date resource for researchers in the field. We analysed approximately 38 papers investigating sleep stage classification employing various machine learning techniques integrated with combined signals. In this study, we describe the models proposed in the existing literature for sleep stage classification, discuss their limitations, and identify potential areas for future research.

Sleep disorders include a huge variety of diseases and their diagnosis very often requires complex and expensive biosignal recordings (polysomnography). Special interest lies in the visual classification of sleep stages in 30 sec windows (epochs) out of this biosignal data, requiring sufficient man-power and experience as well as time. Using the FFT of the heart rate signal to extract the LF/HF frequency power ratio as well as relative peak frequency power within the HF band in combination with the variability of peak frequency power in the HF band we were able to classify sleep stages with an automated algorithm. As signal source we used the ECG signal recorded during the night included in polysomnography recordings from the SIESTA database. The visual 30 sec epoch classification of sleep stages is stored together with the signal data, thereby allowing the comparison of visual classifications to those of the algorithm. While the absolute accuracy of correctly assigned epochs was only 57.5%, our approach can provide a rough overview of the distribution of sleep stages as the clinically relevant result. Therefore, this study represents a first step towards the sleep stage classification from the heart rate signal.