Sleep Stages Research Articles

Insights from the 2nd China intelligent sleep staging competition.

Artificial intelligence (AI) is quickly advancing in the field of sleep medicine, which bodes well for the potential of actual clinical use. In this study, an analysis of the 2nd China Intelligent Sleep Staging Competition was conducted to gain insights into the general level and constraints of AI-assisted sleep staging in China. The outcomes of 10 teams from the children's track and 13 teams from the adult track were investigated in this study. The analysis included overall performance, differences between five different sleep stages, variations across subjects, and performance during stage transitions. The adult track's accuracy peaked at 80.46%, while the children's track's accuracy peaked at 88.96%. On average, accuracy rates stood at 71.43% for children and 68.40% for adults. All results were produced within a mere 5-min timeframe. The N1 stage was prone to misclassification as W, N2, and R stages. In the adult track, significant differences were apparent among subjects (p < 0.05), whereas in the children's track, such differences were not observed. Nonetheless, both tracks experienced a performance decline during stage transitions. The computational speed of AI is remarkably fast, simultaneously holding the potential to surpass the accuracy of physicians. Improving the machine learning model's classification of the N1 stage and transitional periods between stages, along with bolstering its robustness to individual subject variations, is imperative for maximizing its ability in assisting clinical scoring.

Validation of spectral sleep scoring with polysomnography using forehead EEG device

IntroductionVisual scoring of sleep electroencephalography (EEG) has long been considered the gold standard for sleep staging. However, it has several drawbacks, including high cost, time-intensiveness, vulnerability to human variability, discomfort to patients, lack of visualization to validate the hypnogram, and no acknowledgment of differences between delta and slow oscillation deep sleep. This report highlights a spectral scoring approach that addresses all these shortcomings of visual scoring. Past algorithms have used spectral information to help classify traditional visual stages. The current method used the clearly visible spectral patterns to develop new spectral stages, which are similar to but not the same as visual stages. Importantly, spectral scoring delivers both a hypnogram and a whole-night spectrogram, which can be visually inspected to ensure accurate scoring.MethodsThis study compared traditional visual scoring of 32-channel polysomnography with forehead-only spectral scoring from an EEG patch worn concurrently. The PSG was visually scored by trained technicians and the forehead patch was scored spectrally. Because non-rapid eye movement (NREM) stage divisions in spectral scoring are not based on visual NREM stages, the agreements are not expected to be as high as other automated sleep scoring algorithms. Rather, they are a guide to understanding spectral stages as they relate to the more widely understood visual stages and to emphasize reasons for the differences.ResultsThe results showed that visual REM was highly recognized as spectral REM (89%). Visual wake was only scored as spectral Wake 47% of the time, partly because of excessive visual scoring of wake during Light and REM sleep. The majority of spectral Light (predominance of spindle power) was scored as N2 (74%), while less N2 was scored as Light (65%), mostly because of incorrect visual staging of Lo Deep sleep due to high-pass filtering. N3 was scored as both Hi Deep (13 Hz power, 42%) and Lo Deep (0–1 Hz power, 39%), constituting a total of 81% of N3.DiscussionThe results show that spectral scoring better identifies clinically relevant physiology at a substantially lower cost and in a more reproducible fashion than visual scoring, supporting further work exploring its use in clinical and research settings.

Two Cases of Sleep-Related Dissociative Disorder with Episodes of Nocturnal Eating

AbstractSleep-related dissociative disorder (SRDD) is a female-predominant psychiatric parasomnia that was first identified as a condition that mimics sleepwalking in 1989, and was included in the International Classification of Sleep Disorders, 2nd edition, in 2005, with a subsequent expanding literature of case series and case reports. The objective hallmark of SRDD, found in about half of the reported cases, is sustained electroencephalogram (EEG) wakefulness during dissociative episodes emerging during wake-sleep transitions or after awakenings from light non-rapid eye movement (NREM) sleep or rapid eye movement (REM) sleep. Herein are reported two additional cases of SRDD in two female patients aged 53 and 40 years with prominent histories of multimodal abuse (typical of SRDD), with childhood emotional and food deprivation abuse in Case 1, and childhood emotional, sexual, and physical abuse in Case 2. Both patients were affected by “sleep phobia” and had recurrent nocturnal eating episodes. Major findings from the cumulative literature on SRDD are reinforced by these cases, with additional findings being described, particularly nocturnal eating behaviors and priming/triggering factors.

Generalized sleep decoding with basal ganglia signals in multiple movement disorders

Sleep disturbances profoundly affect the quality of life in individuals with neurological disorders. Closed-loop deep brain stimulation (DBS) holds promise for alleviating sleep symptoms, however, this technique necessitates automated sleep stage decoding from intracranial signals. We leveraged overnight data from 121 patients with movement disorders (Parkinson’s disease, Essential Tremor, Dystonia, Essential Tremor, Huntington’s disease, and Tourette’s syndrome) in whom synchronized polysomnograms and basal ganglia local field potentials were recorded, to develop a generalized, multi-class, sleep specific decoder – BGOOSE. This generalized model achieved 85% average accuracy across patients and across disease conditions, even in the presence of recordings from different basal ganglia targets. Furthermore, we also investigated the role of electrocorticography on decoding performances and proposed an optimal decoding map, which was shown to facilitate channel selection for optimal model performances. BGOOSE emerges as a powerful tool for generalized sleep decoding, offering exciting potentials for the precision stimulation delivery of DBS and better management of sleep disturbances in movement disorders.

Neurophysiological features of STN LFP underlying sleep fragmentation in Parkinson’s disease

BackgroundSleep fragmentation is a persistent problem throughout the course of Parkinson’s disease (PD). However, the related neurophysiological patterns and the underlying mechanisms remained unclear.MethodWe recorded subthalamic nucleus (STN) local field...

Older adults at greater risk for Alzheimer’s disease show stronger associations between sleep apnea severity in REM sleep and verbal memory

BackgroundObstructive sleep apnea (OSA) increases risk for cognitive decline and Alzheimer’s disease (AD). While the underlying mechanisms remain unclear, hypoxemia during OSA has been implicated in cognitive impairment. OSA during rapid eye movement (REM) sleep is usually more severe than in non-rapid eye movement (NREM) sleep, but the relative effect of oxyhemoglobin desaturation during REM versus NREM sleep on memory is not completely characterized. Here, we examined the impact of OSA, as well as the moderating effects of AD risk factors, on verbal memory in a sample of middle-aged and older adults with heightened AD risk.MethodsEighty-one adults (mean age:61.7 ± 6.0 years, 62% females, 32% apolipoprotein E ε4 allele (APOE4) carriers, and 70% with parental history of AD) underwent clinical polysomnography including assessment of OSA. OSA features were derived in total, NREM, and REM sleep. REM-NREM ratios of OSA features were also calculated. Verbal memory was assessed with the Rey Auditory Verbal Learning Test (RAVLT). Multiple regression models evaluated the relationships between OSA features and RAVLT scores while adjusting for sex, age, time between assessments, education years, body mass index (BMI), and APOE4 status or parental history of AD. The significant main effects of OSA features on RAVLT performance and the moderating effects of AD risk factors (i.e., sex, age, APOE4 status, and parental history of AD) were examined.ResultsApnea–hypopnea index (AHI), respiratory disturbance index (RDI), and oxyhemoglobin desaturation index (ODI) during REM sleep were negatively associated with RAVLT total learning and long-delay recall. Further, greater REM-NREM ratios of AHI, RDI, and ODI (i.e., more events in REM than NREM) were related to worse total learning and recall. We found specifically that the negative association between REM ODI and total learning was driven by adults 60 + years old. In addition, the negative relationships between REM-NREM ODI ratio and total learning, and REM-NREM RDI ratio and long-delay recall were driven by APOE4 carriers.ConclusionGreater OSA severity, particularly during REM sleep, negatively affects verbal memory, especially for people with greater AD risk. These findings underscore the potential importance of proactive screening and treatment of REM OSA even if overall AHI appears low.

Sleep stage continuity is associated with objective daytime sleepiness in patients with suspected obstructive sleep apnea.

Excessive daytime sleepiness (EDS) in patients with obstructive sleep apnea (OSA) is poorly explained by standard clinical sleep architecture metrics. We hypothesized that reduced sleep stage continuity mediates this connection independently from standard sleep architecture metrics. 1,907 patients with suspected OSA with daytime sleepiness complaints underwent in-lab diagnostic polysomnography and next-day Multiple Sleep Latency Test (MSLT). Sleep architecture was evaluated with novel sleep-stage continuity quantifications (mean sleep stage duration and probability of remaining in each sleep stage), and conventional metrics (total N1, N2, N3 and REM times; and sleep onset latency). Multivariate analyses were utilized to identify variables associated with moderate EDS (5 ≤ mean daytime sleep latency (MSL) ≤ 10 minutes) and severe EDS (MSL < 5 minutes). Compared to those without EDS, participants with severe EDS had lower N3 sleep continuity (mean N3 period duration 10.4 vs 13.7 minutes, p<0.05), less N3 time (53.8 vs 76.5 minutes, p<0.05); greater total sleep time (374.0 vs 352.5 minutes, p<0.05) and greater N2 time (227.5 vs 186.8 minutes, p<0.05). After adjusting for standard sleep architecture metrics using multivariate logistic regression, decreased mean wake and N3 period duration, and the decreased probability of remaining in N2 and N3 sleep remained significantly associated with severe EDS, while the decreased probability of remaining in wake and N2 sleep were associated with moderate EDS. Patients with OSA with EDS experience lower sleep continuity, noticeable especially during N3 sleep and wake. Sleep-stage continuity quantifications assist in characterizing the sleep architecture and are associated with objective daytime sleepiness highlighting the need for more detailed evaluations of sleep quality.

Shared EEG correlates between non-REM parasomnia experiences and dreams

Sleepwalking and related parasomnias result from incomplete awakenings out of non-rapid eye movement sleep. Behavioral episodes can occur without consciousness or recollection, or in relation to dream-like experiences. To understand what accounts for these differences in consciousness and recall, here we recorded parasomnia episodes with high-density electroencephalography (EEG) and interviewed participants immediately afterward about their experiences. Compared to reports of no experience (19%), reports of conscious experience (56%) were preceded by high-amplitude EEG slow waves in anterior cortical regions and activation of posterior cortical regions, similar to previously described EEG correlates of dreaming. Recall of the content of the experience (56%), compared to no recall (25%), was associated with higher EEG activation in the right medial temporal region before movement onset. Our work suggests that the EEG correlates of parasomnia experiences are similar to those reported for dreams and may thus reflect core physiological processes involved in sleep consciousness.

Retuning of hippocampal representations during sleep.

Hippocampal representations that underlie spatial memory undergo continuous refinement following formation1. Here, to track the spatial tuning of neurons dynamically during offline states, we used a new Bayesian learning approach based on the spike-triggered average decoded position in ensemble recordings from freely moving rats. Measuring these tunings, we found spatial representations within hippocampal sharp-wave ripples that were stable for hours during sleep and were strongly aligned with place fields initially observed during maze exploration. These representations were explained by a combination of factors that included preconfigured structure before maze exposure and representations that emerged during θ-oscillations and awake sharp-wave ripples whileon the maze, revealing the contribution of these events in forming ensembles. Strikingly, the ripple representations during sleep predicted the future place fields of neurons during re-exposure to the maze, even when those fields deviated from previous place preferences. By contrast, we observed tunings with poor alignment to maze place fields during sleep and rest before maze exposure and in the later stages of sleep. In sum, the new decoding approach allowed us to infer and characterize the stability and retuning of place fields during offline periods, revealing the rapid emergence of representations following new exploration and the role of sleep in the representational dynamics of the hippocampus.

Age-Related Changes in Sleep and Its Implications for Cognitive Decline in Aging Persons With Schizophrenia: A Critical Review.

Cognitive impairment is a core feature of schizophrenia that worsens with aging and interferes with quality of life. Recent work identifies sleep as an actionable target to alleviate cognitive deficits. Cardinal non-rapid eye movement (NREM) sleep oscillations such as sleep spindles and slow oscillations are critical for cognition. People living with schizophrenia (PLWS) and their first-degree relatives have a specific reduction in sleep spindles and an abnormality in their temporal coordination with slow oscillations that predict impaired memory consolidation. While NREM oscillatory activity is reduced in typical aging, it is not known how further disruption in these oscillations contributes to cognitive decline in older PLWS. Another understudied risk factor for cognitive deficits among older PLWS is obstructive sleep apnea (OSA) which may contribute to cognitive decline. We conducted a narrative review to examine the published literature on aging, OSA, and NREM sleep oscillations in PLWS. Spindles are propagated via thalamocortical feedback loops, and this circuitry shows abnormal hyperconnectivity in schizophrenia as revealed by structural and functional MRI studies. While the risk and severity of OSA increase with age, older PLWS are particularly vulnerable to OSA-related cognitive deficits because OSA is often underdiagnosed and undertreated, and OSA adds further damage to the circuitry that generates NREM sleep oscillations. We highlight the critical need to study NREM sleep in older PWLS and propose that identifying and treating OSA in older PLWS will provide an avenue to potentially mitigate and prevent cognitive decline.

Workflow for the unsupervised clustering of sleep stages identifies light and deep sleep in electrophysiological recordings in mice

BackgroundSleep physiology plays a critical role in brain development and aging. Accurate sleep staging, which categorizes different sleep states, is fundamental for sleep physiology studies. Traditional methods for sleep staging rely on manual, rule-based scoring techniques, which limit their accuracy and adaptability. New methodWe describe, test and challenge a workflow for unsupervised clustering of sleep states (WUCSS) in rodents, which uses accelerometer and electrophysiological data to classify different sleep states. WUCSS utilizes unsupervised clustering to identify sleep states using six features, extracted from 4-second epochs. ResultsWe gathered high-quality EEG recordings combined with accelerometer data in diverse transgenic mouse lines (male ApoE3 versus ApoE4 knockin; male CNTNAP2 KO versus wildtype littermates). WUCSS showed high recall, precision, and F1-score against manual scoring on awake, NREM, and REM sleep states. Within NREM, WUCSS consistently identified two additional clusters that qualify as deep and light sleep states. Comparison with existing methodsThe ability of WUCSS to discriminate between deep and light sleep enhanced the precision and comprehensiveness of the current mouse sleep physiology studies. This differentiation led to the discovery of an additional sleep phenotype, notably in CNTNAP2 KO mice, showcasing the method's superiority over traditional scoring methods. ConclusionsWUCSS, with its unsupervised approach and classification of deep and light sleep states, provides an unbiased opportunity for researchers to enhance their understanding of sleep physiology. Its high accuracy, adaptability, and ability to save time and resources make it a valuable tool for improving sleep staging in both clinical and preclinical research.

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 multi-task learning model using RR intervals and respiratory effort to assess sleep disordered breathing

BackgroundSleep-disordered breathing (SDB) affects a significant portion of the population. As such, there is a need for accessible and affordable assessment methods for diagnosis but also case-finding and long-term follow-up. Research has focused on exploiting cardiac and respiratory signals to extract proxy measures for sleep combined with SDB event detection. We introduce a novel multi-task model combining cardiac activity and respiratory effort to perform sleep–wake classification and SDB event detection in order to automatically estimate the apnea–hypopnea index (AHI) as severity indicator.MethodsThe proposed multi-task model utilized both convolutional and recurrent neural networks and was formed by a shared part for common feature extraction, a task-specific part for sleep–wake classification, and a task-specific part for SDB event detection. The model was trained with RR intervals derived from electrocardiogram and respiratory effort signals. To assess performance, overnight polysomnography (PSG) recordings from 198 patients with varying degree of SDB were included, with manually annotated sleep stages and SDB events.ResultsWe achieved a Cohen’s kappa of 0.70 in the sleep–wake classification task, corresponding to a Spearman’s correlation coefficient (R) of 0.830 between the estimated total sleep time (TST) and the TST obtained from PSG-based sleep scoring. Combining the sleep–wake classification and SDB detection results of the multi-task model, we obtained an R of 0.891 between the estimated and the reference AHI. For severity classification of SBD groups based on AHI, a Cohen’s kappa of 0.58 was achieved. The multi-task model performed better than a single-task model proposed in a previous study for AHI estimation, in particular for patients with a lower sleep efficiency (R of 0.861 with the multi-task model and R of 0.746 with single-task model with subjects having sleep efficiency < 60%).ConclusionAssisted with automatic sleep–wake classification, our multi-task model demonstrated proficiency in estimating AHI and assessing SDB severity based on AHI in a fully automatic manner using RR intervals and respiratory effort. This shows the potential for improving SDB screening with unobtrusive sensors also for subjects with low sleep efficiency without adding additional sensors for sleep–wake detection.

Pengaruh Endorphine Massage Terhadap Kualitas Tidur Ibu Hamil Usia Kehamilan 37- 40 Minggu

The decrease sleep duration during pregnancy increases in the third trimester. Giving non-pharmacological therapies that can be done to help improve sleep quality is to do an endorphine massage. The aim of this study was to determine the effect of endorphine massage on the sleep quality of pregnant women 37-40 weeks in the Working Area of the Mojo Health Center, Kediri Regency. The method of this research is Pre-Experiment with One Group Pretest-Posttest Design without a control group, using a total sampling technique with a total sample of 30 respondents. Endorphine massage intervention will be carried out 2 times in a row with the duration of 15-20 minutes each session. Data collection used the questionnaire PSQI, analysis used Mc Nemar test obtained P-Value 1 &gt; 0.05, it can be concluded that statistically, there is no effect of endorphine massage on the sleep quality of pregnant women 37-40 weeks in the Working Area of the Mojo Health Center, Kediri Regency. Poor sleep quality in pregnant women due to frequent awakening from sleep. This makes it difficult for pregnant women to achieve deep sleep. So it is necessary to do endorphine massage regularly by paying attention to disturbing factors, so that pregnant women can experience better quality sleep.

SleepBoost: a multi-level tree-based ensemble model for automatic sleep stage classification.

Neurodegenerative diseases often exhibit a strong link with sleep disruption, highlighting the importance of effective sleep stage monitoring. In this light, automatic sleep stage classification (ASSC) plays a pivotal role, now more streamlined than ever due to the advancements in deep learning (DL). However, the opaque nature of DL models can be a barrier in their clinical adoption, due to trust concerns among medical practitioners. To bridge this gap, we introduce SleepBoost, a transparent multi-level tree-based ensemble model specifically designed for ASSC. Our approach includes a crafted feature engineering block (FEB) that extracts 41 time and frequency domain features, out of which 23 are selected based on their high mutual information score (> 0.23). Uniquely, SleepBoost integrates three fundamental linear models into a cohesive multi-level tree structure, further enhanced by a novel reward-based adaptive weight allocation mechanism. Tested on the Sleep-EDF-20 dataset, SleepBoost demonstrates superior performance with an accuracy of 86.3%, F1-score of 80.9%, and Cohen kappa score of 0.807, outperforming leading DL models in ASSC. An ablation study underscores the critical role of our selective feature extraction in enhancing model accuracy and interpretability, crucial for clinical settings. This innovative approach not only offers a more transparent alternative to traditional DL models but also extends potential implications for monitoring and understanding sleep patterns in the context of neurodegenerative disorders. The open-source availability of SleepBoost's implementation at https://github.com/akibzaman/SleepBoost can further facilitate its accessibility and potential for widespread clinical adoption.

Jerks are Useful: Extracting pulse rate from wrist-placed accelerometry jerk during sleep in children.

Evaluate wrist-placed accelerometry predicted heartrate compared to electrocardiogram (ECG) heartrate in children during sleep. Children (n=82, 61% male, 43.9% Black) wore a wrist-placed Apple Watch Series 7 (AWS7) and ActiGraph GT9X during a polysomnogram. 3-Axis accelerometry data was extracted from AWS7 and the GT9X. Accelerometry heartrate estimates were derived from jerk (the rate of acceleration change), computed using the peak magnitude frequency in short time Fourier Transforms of Hilbert transformed jerk computed from acceleration magnitude. Heartrates from ECG traces were estimated from R-R intervals using R-pulse detection. Lin's Concordance Correlation Coefficient (CCC), mean absolute error (MAE) and mean absolute percent error (MAPE) assessed agreement with ECG estimated heartrate. Secondary analyses explored agreement by polysomnography sleep stage and a signal quality metric. The developed scripts are available on Github. For the GT9X, CCC was poor at -0.11 and MAE and MAPE were high at 16.8 (SD=14.2) beats/minute and 20.4% (SD=18.5%). For AWS7, CCC was moderate at 0.61 while MAE and MAPE were lower at 6.4 (SD=9.9) beats/minute and 7.3% (SD=10.3%). Accelerometry estimated heartrate for AWS7 was more closely related to ECG heartrate during N2, N3 and REM sleep than lights on, wake, and N1 and when signal quality was high. These patterns were not evident for the GT9X. Raw accelerometry data extracted from AWS7, but not the GT9X, can be used to estimate heartrate in children while they sleep. Future work is needed to explore the sources (i.e., hardware, software, etc.) of the GT9X's poor performance.

Microglial TNFα controls daily changes in synaptic GABAARs and sleep slow waves.

Microglia sense the changes in their environment. How microglia actively translate these changes into suitable cues to adapt brain physiology is unknown. We reveal an activity-dependent regulation of cortical inhibitory synapses by microglia, driven by purinergic signaling acting on P2RX7 and mediated by microglia-derived TNFα. We demonstrate that sleep induces microglia-dependent synaptic enrichment of GABAARs in a manner dependent on microglial TNFα and P2RX7. We further show that microglia-specific depletion of TNFα alters slow waves during NREM sleep and blunt memory consolidation in sleep-dependent learning tasks. Together, our results reveal that microglia orchestrate sleep-intrinsic plasticity of synaptic GABAARs, sculpt sleep slow waves, and support memory consolidation.

Digital sleep clinic: assessing efficacy of continuous positive airway pressure through sleep staging via connected devices: a study protocol

BackgroundObstructive sleep apnea (OSA) is a multisystemic chronic disease with disabling symptoms, cardiometabolic comorbidities and reduction in physical activity. Continuous positive airway pressure (CPAP) is the standard treatment for OSA. Only a few studies have characterized trajectories of sleep parameters upon initiation of CPAP and these are limited to one or two nights of polysomnographic recording in a sleep laboratory. This is due to the cost of carrying out these studies and poor tolerance by patients of multiple nights of polysomnographic recordings. No study has characterized sleep over multiple nights before and after CPAP initiation, assessing the multidimensional efficacy of CPAP on patient reported outcomes, objective and subjective sleep quality, oximetry, glucose control and physical activity.New digital technologies enable overnight sleep studies over several nights in the patient’s home, with a reliability of sleep characterization equivalent to polysomnographic recording.The primary aim of this study is to investigate objective slow wave sleep (SWS or N3) quality before CPAP and during the first month of the treatment.Secondary objectives are to assess changes in the following parameters before CPAP and during the first month of the treatment: other objective sleep parameters and sleep stages evolution (W, N1, N2 and REM), nocturnal oxygen desaturations, 24-h blood glucose profile, daily physical activity (the daily steps count), and patient reported outcomes.MethodsSeventy patients prescribed CPAP for OSA will be recruited at Grenoble Alpes University Hospital (France) and monitored for 5 weeks using validated innovative wearable connected devices (the Dreem 3 headband, a pedometer, an oximeter, and a continuous glucose sensor) enabling them to track their own sleep and physiological parameters at home before and after CPAP initiation.DiscussionBy pooling data from the CPAP telemonitoring and other connected devices we should be able to follow the multidimensional trajectories of patients after the initiation of CPAP. This will enable us to determine whether objective changes in sleep parameters in the first few weeks of CPAP treatment are associated with improvements in daytime sleepiness, quality of life, treatment adherence, glucose control and physical activity. The data will provide integrated markers of treatment efficacy and will allow adapted personalized management of OSA in the short and long-term.Trial registrationClinicaltrials (NCT05197855).

Single-channel EOG sleep staging on a heterogeneous cohort of subjects with sleep disorders

Objective. Sleep staging based on full polysomnography is the gold standard in the diagnosis of many sleep disorders. It is however costly, complex, and obtrusive due to the use of multiple electrodes. Automatic sleep staging based on single-channel electro-oculography (EOG) is a promising alternative, requiring fewer electrodes which could be self-applied below the hairline. EOG sleep staging algorithms are however yet to be validated in clinical populations with sleep disorders. Approach. We utilized the SOMNIA dataset, comprising 774 recordings from subjects with various sleep disorders, including insomnia, sleep-disordered breathing, hypersomnolence, circadian rhythm disorders, parasomnias, and movement disorders. The recordings were divided into train (574), validation (100), and test (100) groups. We trained a neural network that integrated transformers within a U-Net backbone. This design facilitated learning of arbitrary-distance temporal relationships within and between the EOG and hypnogram. Main results. For 5-class sleep staging, we achieved median accuracies of 85.0% and 85.2% and Cohen’s kappas of 0.781 and 0.796 for left and right EOG, respectively. The performance using the right EOG was significantly better than using the left EOG, possibly because in the recommended AASM setup, this electrode is located closer to the scalp. The proposed model is robust to the presence of a variety of sleep disorders, displaying no significant difference in performance for subjects with a certain sleep disorder compared to those without. Significance. The results show that accurate sleep staging using single-channel EOG can be done reliably for subjects with a variety of sleep disorders.

Sleep staging algorithm based on smartwatch sensors for healthy and sleep apnea populations

ObjectiveSleep stages can provide valuable insights into an individual’s sleep quality. By leveraging movement and heart rate data collected by modern smartwatches, it is possible to enable the sleep staging feature and enhance users’ understanding about their sleep and health conditions. MethodIn this paper, we present and validate a recurrent neural network based model with 23 input features extracted from accelerometer and photoplethysmography sensors data for both healthy and sleep apnea populations. We designed a lightweight and fast solution to enable the prediction of sleep stages for each 30-seconds epoch. This solution was developed using a large dataset of 1,522 night recordings collected from a highly heterogeneous population and different versions of Samsung smartwatch. ResultsIn the classification of four sleep stages (wake, light, deep, and rapid eye movements sleep), the proposed solution achieved 71.6% of balanced accuracy and a Cohen’s kappa of 0.56 in a test set with 586 recordings. ConclusionThe results presented in this paper validate our proposal as a competitive wearable solution for sleep staging. Additionally, the use of a large and diverse data set contributes to the robustness of our solution, and corroborates the validation of algorithm’s performance. Some additional analysis performed for healthy and sleep apnea population demonstrated that algorithm’s performance has low correlation with demographic variables.