scholarly journals Automatic Wake-Sleep Stages Classification using Electroencephalogram Instantaneous Frequency and Envelope Tracking

2020 ◽  
Author(s):  
Mahdi Rahbar Alam ◽  
Reza Sameni
Keyword(s):  
Instantaneous Frequency ◽  
Sleep Stage ◽  
Analytic Form ◽  
Dominant Frequency ◽  
Research Field ◽  
Frequency Resolution ◽  
Sleep Stages ◽  
Eeg Spectrum ◽  
Robust Model ◽  
The Rich

AbstractBackgroundThe study of cerebral activity during sleep using the electroencephalograph (EEG) is a major research field in neuroscience. Despite the rich literature in this field, the automatic and accurate categorization of wake-sleep stages remains an open problem.New MethodA robust model-based Kalman filtering scheme is proposed for tracking the poles of a second order time-varying autoregressive model fitted over the EEG acquired during different wake/sleep stages. The pole angle/phase is regarded as the dominant frequency of the EEG spectrum (known as the instantaneous frequency in literature). The frequency resolution is improved by splitting the wide frequency band to subbands corresponding to well-known brain rhythms. Using recent findings in field of EEG phase/frequency tracking, the instantaneous envelope of the narrow-band signal’s analytic form is also tracked as a complementary feature.ResultsThe minimal set of instantaneous frequency and envelope features is employed in three classification schemes, using training labels from R&k and AASM sleep scoring standards. The LDA classifier resulted in the highest performance using the proposed feature set.Comparison with Existing MethodsThe proposed method resulted in a higher mean decoding accuracy and a lower standard deviation on the entire dataset, as compared with state-of-the-art techniques.ConclusionsThe accurate tracking of the instantaneous frequency and envelope are highly informative for sleep stage scoring. The proposed method is shown to have additional applications, including the prediction of wake-sleep transition, which can be used for drowsiness detection from the EEG.

Sleep Stage Classification Based on EEG Signals by Using Improved Hilbert-Huang Transform

2011 ◽  
Vol 138-139 ◽  
pp. 1096-1101
Author(s):  
Xue Li Shen ◽  
Ying Le Fan
Keyword(s):  
Sleep Stage ◽  
Frequency Resolution ◽  
Optimal Decision ◽  
Spectral Energy ◽  
Sleep Stages ◽  
Sleep Staging ◽  
Intrinsic Mode Functions ◽  
Eeg Signals ◽  
Hilbert Huang Transform ◽  
Time Frequency

Research on automatic sleep staging based on EEG signals has a significant meaning for objective evaluation of sleep quality. An improved Hilbert-Huang transform method was applied to time-frequency analysis of non-stable EEG signals for the sleep staging in this paper. In order to settle the frequency overlapping problem of intrinsic mode function obtained from traditional HHT, wavelet package transform was introduced to bandwidth refinement of EEG before the empirical mode decomposition was conducted. This method improved the time-frequency resolution effectively. Then the intrinsic mode functions and their marginal spectrums would be calculated. Six common spectrum energies (or spectral energy ratios) were selected as characteristic parameters. Finally, a probabilistic nearest neighbor method for statistical pattern recognition was applied to optimal decision. The experiment data was from the Sleep-EDF database of MIT-BIH. The classification results showed that the automatic sleep staging decisions based on this method conformed roughly with the manual staging results and were better than those obtained from traditional HHT obviously. Therefore, the method in this paper could be applied to extract features of sleep stages and provided necessary dependence for automatic sleep staging.

scholarly journals Classification of Brainwaves for Sleep Stages by High-Dimensional FFT Features from EEG Signals

2020 ◽  
Vol 10 (5) ◽  
pp. 1797 ◽  
Author(s):  
Mera Kartika Delimayanti ◽  
Bedy Purnama ◽  
Ngoc Giang Nguyen ◽  
Mohammad Reza Faisal ◽  
Kunti Robiatul Mahmudah ◽  
...  
Keyword(s):  
Machine Learning ◽  
Sleep Stage ◽  
Machine Learning Algorithms ◽  
High Dimensional ◽  
Sleep Stages ◽  
Eeg Signals ◽  
Stage Classification ◽  
Sleep Stage Classification ◽  
Low Dimensional

Manual classification of sleep stage is a time-consuming but necessary step in the diagnosis and treatment of sleep disorders, and its automation has been an area of active study. The previous works have shown that low dimensional fast Fourier transform (FFT) features and many machine learning algorithms have been applied. In this paper, we demonstrate utilization of features extracted from EEG signals via FFT to improve the performance of automated sleep stage classification through machine learning methods. Unlike previous works using FFT, we incorporated thousands of FFT features in order to classify the sleep stages into 2–6 classes. Using the expanded version of Sleep-EDF dataset with 61 recordings, our method outperformed other state-of-the art methods. This result indicates that high dimensional FFT features in combination with a simple feature selection is effective for the improvement of automated sleep stage classification.

scholarly journals Capturing the Development of Internal Representations in a High-Performing Deep Network for Sleep Stage Classification

2021 ◽  
Vol 2 (4) ◽  
Author(s):  
Sarun Paisarnsrisomsuk ◽  
Carolina Ruiz ◽  
Sergio A. Alvarez
Keyword(s):  
Input Data ◽  
Single Channel ◽  
Sleep Stage ◽  
Quantitative Measure ◽  
Classification Performance ◽  
Sleep Stages ◽  
Layer Depth ◽  
Human Sleep ◽  
Stage Classification ◽  
Sleep Stage Classification

AbstractDeep neural networks can provide accurate automated classification of human sleep signals into sleep stages that enables more effective diagnosis and treatment of sleep disorders. We develop a deep convolutional neural network (CNN) that attains state-of-the-art sleep stage classification performance on input data consisting of human sleep EEG and EOG signals. Nested cross-validation is used for optimal model selection and reliable estimation of out-of-sample classification performance. The resulting network attains a classification accuracy of $$84.50 \pm 0.13\%$$ 84.50 ± 0.13 % ; its performance exceeds human expert inter-scorer agreement, even on single-channel EEG input data, therefore providing more objective and consistent labeling than human experts demonstrate as a group. We focus on analyzing the learned internal data representations of our network, with the aim of understanding the development of class differentiation ability across the layers of processing units, as a function of layer depth. We approach this problem visually, using t-Stochastic Neighbor Embedding (t-SNE), and propose a pooling variant of Centered Kernel Alignment (CKA) that provides an objective quantitative measure of the development of sleep stage specialization and differentiation with layer depth. The results reveal a monotonic progression of both of these sleep stage modeling abilities as layer depth increases.

Sleep Stage Assessment Using Power Spectral Indices of Heart Rate Variability With a Simple Algorithm

2012 ◽  
Vol 15 (3) ◽  
pp. 264-272 ◽  
Author(s):  
Keiko Tanida ◽  
Masashi Shibata ◽  
Margaret M. Heitkemper
Keyword(s):  
Heart Rate ◽  
Heart Rate Variability ◽  
Sleep Stage ◽  
Sleep Onset ◽  
Low Frequency ◽  
Simple Algorithm ◽  
Sleep Stages ◽  
Spectral Indices ◽  
Current Form ◽  
Power Spectral

Clinical researchers do not typically assess sleep with polysomnography (PSG) but rather with observation. However, methods relying on observation have limited reliability and are not suitable for assessing sleep depth and cycles. The purpose of this methodological study was to compare a sleep analysis method based on power spectral indices of heart rate variability (HRV) data to PSG. PSG and electrocardiography data were collected synchronously from 10 healthy women (ages 20–61 years) over 23 nights in a laboratory setting. HRV was analyzed for each 60-s epoch and calculated at 3 frequency band powers (very low frequency [VLF]-hi: 0.016–0.04 Hz; low frequency [LF]: 0.04–0.15 Hz; and high frequency [HF]: 0.15–0.4 Hz). Using HF/(VLF-hi + LF + HF) value, VLF-hi, and heart rate (HR) as indices, an algorithm to categorize sleep into 3 states (shallow sleep corresponding to Stages 1 & 2, deep sleep corresponding to Stages 3 & 4, and rapid eye movement [REM] sleep) was created. Movement epochs and time of sleep onset and wake-up were determined using VLF-hi and HR. The minute-by-minute agreement rate with the sleep stages as identified by PSG and HRV data ranged from 32 to 72% with an average of 56%. Longer wake after sleep onset (WASO) resulted in lower agreement rates. The mean differences between the 2 methods were 2 min for the time of sleep onset and 6 min for the time of wake-up. These results indicate that distinguishing WASO from shallow sleep segments is difficult using this HRV method. The algorithm's usefulness is thus limited in its current form, and it requires additional modification.

scholarly journals Light Sleep Detection based on Surface Electromyography Signals for Nap Monitoring

2022 ◽  
Vol 16 ◽  
pp. 140-145
Author(s):  
Wachiraporn Aiamklin ◽  
Yutana Jewajinda ◽  
Yunyong Punsawad
Keyword(s):  
Surface Electromyography ◽  
Sleep Stage ◽  
Low Cost ◽  
Trapezius Muscle ◽  
Sleep Stages ◽  
Upper Trapezius ◽  
Emg Signal ◽  
Power Frequency ◽  
Upper Trapezius Muscle ◽  
Waking Up

This paper proposes the development of automatic sleep stage detection by using physiological signals. We aim to develop an application to assist drivers after drowsiness or fatigue detection by a commercial driver vigilance system. The proposed method used a low-cost surface electromyography (EMG) device for sleep stage detection. We investigate skeletal muscle location and EMG features from sleep stage 2 to provide an EMG-based nap monitoring system. The results showed that using only one channel of a bipolar EMG signal from an upper trapezius muscle with median power frequency can achieve 84% accuracy. We implement a MyoWare muscle sensor into the proposed nap monitoring device. The results showed that the proposed system is feasible for detecting sleep stages and waking up the napper. A combination of EMG and electroencephalogram (EEG) signals might be yield a high system performance for nap monitoring and alarm system. We will prototype a portable device to connect the application to a smartphone and test with a target group, such as truck drivers and physicians.

nCPAP Treatment of Obstructive Sleep Apnea Increases Slow Wave Sleep in Prefrontal EEG

2007 ◽  
Vol 38 (3) ◽  
pp. 148-154 ◽  
Author(s):  
Veera Eskelinen ◽  
Toomas Uibu ◽  
Sari-Leena Himanen
Keyword(s):  
Obstructive Sleep Apnea ◽  
Sleep Apnea ◽  
Treatment Period ◽  
Slow Wave ◽  
Slow Wave Sleep ◽  
Sleep Stage ◽  
Sleep Eeg ◽  
Sleep Stages ◽  
Obstructive Sleep ◽  
Sleep Stage Scoring

According to standard sleep stage scoring, sleep EEG is studied from the central area of parietal lobes. However, slow wave sleep (SWS) has been found to be more powerful in frontal areas in healthy subjects. Obstructive sleep apnea syndrome (OSAS) patients often suffer from functional disturbances in prefrontal lobes. We studied the effects of nasal Continuous Positive Airway Pressure (nCPAP) treatment on sleep EEG, and especially on SWS, in left prefrontal and central locations in 12 mild to moderate OSAS patients. Sleep EEG was recorded by polysomnography before treatment and after a 3 month nCPAP treatment period. Recordings were classified into sleep stages. No difference was found in SWS by central sleep stage scoring after the nCPAP treatment period, but in the prefrontal lobe all night S3 sleep stage increased during treatment. Furthermore, prefrontal SWS increased in the second and decreased in the fourth NREM period. There was more SWS in prefrontal areas both before and after nCPAP treatment, and SWS increased significantly more in prefrontal than central areas during treatment. Regarding only central sleep stage scoring, nCPAP treatment did not increase SWS significantly. Frontopolar recording of sleep EEG is useful in addition to central recordings in order to better evaluate the results of nCPAP treatment.

High-throughput visual assessment of sleep stages in mice using machine learning

SLEEP ◽  
2021 ◽  
Author(s):  
Brian Geuther ◽  
Mandy Chen ◽  
Raymond J Galante ◽  
Owen Han ◽  
Jie Lian ◽  
...  
Keyword(s):  
Machine Learning ◽  
Eye Movement ◽  
High Throughput ◽  
Data Augmentation ◽  
Sleep Stage ◽  
Visual Assessment ◽  
Video Data ◽  
Sleep Stages ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep

Abstract Study Objectives Sleep is an important biological process that is perturbed in numerous diseases, and assessment its substages currently requires implantation of electrodes to carry out electroencephalogram/electromyogram (EEG/EMG) analysis. Although accurate, this method comes at a high cost of invasive surgery and experts trained to score EEG/EMG data. Here, we leverage modern computer vision methods to directly classify sleep substages from video data. This bypasses the need for surgery and expert scoring, provides a path to high-throughput studies of sleep in mice. Methods We collected synchronized high-resolution video and EEG/EMG data in 16 male C57BL/6J mice. We extracted features from the video that are time and frequency-based and used the human expert-scored EEG/EMG data to train a visual classifier. We investigated several classifiers and data augmentation methods. Results Our visual sleep classifier proved to be highly accurate in classifying wake, non-rapid eye movement sleep (NREM), and rapid eye movement sleep (REM) states, and achieves an overall accuracy of 0.92 +/- 0.05 (mean +/- SD). We discover and genetically validate video features that correlate with breathing rates, and show low and high variability in NREM and REM sleep, respectively. Finally, we apply our methods to non-invasively detect that sleep stage disturbances induced by amphetamine administration. Conclusions We conclude that machine learning based visual classification of sleep is a viable alternative to EEG/EMG based scoring. Our results will enable non-invasive high-throughput sleep studies and will greatly reduce the barrier to screening mutant mice for abnormalities in sleep.

scholarly journals Sleep, Subcortical Stimulation and Kindling in the Cat

1975 ◽  
Vol 2 (4) ◽  
pp. 447-455 ◽  
Author(s):  
T. Tanaka ◽  
H. Lange ◽  
R. Naquet
Keyword(s):  
Rem Sleep ◽  
Sleep Stage ◽  
Low Frequency ◽  
Mesencephalic Reticular Formation ◽  
Seizure Threshold ◽  
Sleep Stages ◽  
Central Gray Matter ◽  
Spike Discharge ◽  
Amygdaloid Stimulation ◽  
Stimulation Of

SUMMARY:A longitudinal study of the effects of sleep on amygdaloid kindling showed that kindling disrupted normal sleep patterns by reducing REM sleep and increasing awake time. Few interictal spike discharges were observed during the awake stage, while a marked increase in discharge was observed during the light and deep sleep stages. No discharges were observed during REM sleep. During the immediate post-stimulation period the nonstimulated amygdala showed a much higher rate of spike discharge. On the other hand, there was an increase in spike discharge in the stimulated amygdala during natural sleep without preceding amygdaloid stimulation. Amygdaloid stimulation at the generalized seizure threshold during each sleep stage resulted in a generalized convulsion.The influence of subcortical electrical stimulation on kindled amygdaloid convulsions was investigated in a second experiment. Stimulation of the centre median and the caudate nucleus was without effect on kindled convulsions, while stimulation of the mesencephalic reticular formation at high frequency (300 Hz) reduced the latency of onset of kindled generalized convulsions. Stimulation of the nucleus ventralis lateralis of the thalamus at low frequency (10 Hz) prolonged the convulsion latency, and at high current levels blocked the induced convulsion. Stimulation in the central gray matter at low frequency (10 Hz) also blocked kindled amygdaloid convulsions.

scholarly journals Kolmogorov Complexity of Coronary Sinus Atrial Electrograms before Ablation Predicts Termination of Atrial Fibrillation after Pulmonary Vein Isolation

Entropy ◽  
2019 ◽  
Vol 21 (10) ◽  
pp. 970 ◽  
Author(s):  
Stępień ◽  
Kuklik ◽  
Żebrowski ◽  
Sanders ◽  
Derejko ◽  
...  
Keyword(s):  
Atrial Fibrillation ◽  
Catheter Ablation ◽  
Pulmonary Vein ◽  
Pulmonary Vein Isolation ◽  
Coronary Sinus ◽  
Kolmogorov Complexity ◽  
Similarity Index ◽  
Dominant Frequency ◽  
Significant Difference ◽  
The Rich

Atrial fibrillation (AF) is related to a very complex local electrical activity reflected in the rich morphology of intracardiac electrograms. The link between electrogram complexity and efficacy of the catheter ablation is unclear. We test the hypothesis that the Kolmogorov complexity of a single atrial bipolar electrogram recorded during AF within the coronary sinus (CS) at the beginning of the catheter ablation may predict AF termination directly after pulmonary vein isolation (PVI). The study population consisted of 26 patients for whom 30 s baseline electrograms were recorded. In all cases PVI was performed. If AF persisted after PVI, ablation was extended beyond PVs. Kolmogorov complexity estimated by Lempel–Ziv complexity and the block decomposition method was calculated and compared with other measures: Shannon entropy, AF cycle length, dominant frequency, regularity, organization index, electrogram fractionation, sample entropy and wave morphology similarity index. A 5 s window length was chosen as optimal in calculations. There was a significant difference in Kolmogorov complexity between patients with AF termination directly after PVI compared to patients undergoing additional ablation (p < 0.01). No such difference was seen for remaining complexity parameters. Kolmogorov complexity of CS electrograms measured at baseline before PVI can predict self-termination of AF directly after PVI.

scholarly journals 1193 Accuracy of a Commercial Wearable in Detecting Sleep Stages Compared to Polysomnography in Adults: Considering Sleep Classification Methods and Effects of Evening Alcohol Consumption

SLEEP ◽  
2020 ◽  
Vol 43 (Supplement_1) ◽  
pp. A456-A457 ◽  
Author(s):  
L Menghini ◽  
V Alschuler ◽  
S Claudatos ◽  
A Goldstone ◽  
F Baker ◽  
...  
Keyword(s):  
Alcohol Consumption ◽  
Rem Sleep ◽  
Sleep Stage ◽  
Cardiac Activity ◽  
Sleep Stages ◽  
Classification Methods ◽  
Management Platform ◽  
Wake Patterns ◽  
Wake Detection ◽  
Data Management Platform

Abstract Introduction Commercial wearable devices have shown the capability of collecting and processing multisensor information (motion, cardiac activity), claiming to be able to measure sleep-wake patterns and differentiate sleep stages. While using these devices, users should be aware of their accuracy, sources of measurement error and contextual factors that may affect their performance. Here, we evaluated the agreement between Fitbit Charge 2™ and PSG in adults, considering effects of two different sleep classification methods and pre-sleep alcohol consumption. Methods Laboratory-based synchronized recordings of device and PSG data were obtained from 14 healthy adults (42.6±9.7y; 6 women), who slept between one and three nights in the lab, for a total of 27 nights of data. On 10 of these nights, participants consumed alcohol (up to 4 standard drinks) in the 2 hours before bedtime. Device performance relative to PSG was evaluated using epoch-by-epoch and Bland-Altman analyses, with device data obtained from a data-management platform, Fitabase, via two methods one that accounts for short wakes (SW, awakenings that last less than 180s) and one that does not (not-SW). Results SW and not-SW methods were similar in scoring (96.76% agreement across epochs), although the SW method had better accuracy for differentiating “light”, “deep”, and REM sleep; but produced more false positives in wake detection. The device (SW-method) classified epochs of wake, “light” (N1+N2), “deep” (N3) and REM sleep with 56%, 77%, 46%, and 62% sensitivity, respectively. Bland-Altman analysis showed that the device significantly underestimated “light” (~19min) and “deep” (~26min) sleep. Alcohol consumption enhanced PSG-device discrepancies, in particular for REM sleep (p=0.01). Conclusion Our results indicate promising accuracy in sleep-wake and sleep stage identification for this device, particularly when accounting for short wakes, as compared to PSG. Alcohol consumption, as well as other potential confounders that could affect measurement accuracy should be further investigated. Support This study was supported by the National Institute on Alcohol Abuse and Alcoholism (NIAAA) grant R21-AA024841 (IMC and MdZ). The content is solely the responsibility of the authors and does not necessarily represent the official views the National Institutes of Health.

Sign in / Sign up

Export Citation Format

Share Document