CAPSCNet: A novel scattering network for automated identification of phasic cyclic alternating patterns of human sleep using multivariate EEG signals.

The Cyclic Alternating Pattern (CAP) can be considered a physiological marker of sleep instability. The CAP can examine various sleep-related disorders. Certain short events (A and B phases) manifest related to a specific physiological process or pathology during non-rapid eye movement (NREM) sleep. These phases unexpectedly modify EEG oscillations; hence, manual detection is challenging. Therefore, it is highly desirable to have an automated system for detecting the A-phases (AP). Deep convolution neural networks (CNN) have shown high performance in various healthcare applications. A variant of the deep neural network called the Wavelet Scattering Network (WSN) has been used to overcome the specific limitations of CNN, such as the need for a large amount of data to train the model. WSN is an optimized network that can learn features that help discriminate patterns hidden inside signals. Also, WSNs are invariant to local perturbations, making the network significantly more reliable and effective. It can also help improve performance on tasks where data is minimal. In this study, we proposed a novel WSN-based CAPSCNet to automatically detect AP using EEG signals. Seven dataset variants of cyclic alternating pattern (CAP) sleep cohort is employed for this study. Two electroencephalograms (EEG) derivations, namely: C4-A1 and F4-C4, are used to develop the CAPSCNet. The model is examined using healthy subjects and patients tormented by six different sleep disorders, namely: sleep-disordered breathing (SDB), insomnia, nocturnal frontal lobe epilepsy (NFLE), narcolepsy, periodic leg movement disorder (PLM) and rapid eye movement behavior disorder (RBD) subjects. Several different machine-learning algorithms were used to classify the features obtained from the WSN. The proposed CAPSCNet has achieved the highest average classification accuracy of 83.4% using a trilayered neural network classifier for the healthy data variant. The proposed CAPSCNet is efficient and computationally faster.

Automated sleep classification with chronic neural implants in freely behaving canines.

Objective.Long-term intracranial electroencephalography (iEEG) in freely behaving animals provides valuable electrophysiological information and when correlated with animal behavior is useful for investigating brain function.Approach.Here we develop and validate an automated iEEG-based sleep-wake classifier for canines using expert sleep labels derived from simultaneous video, accelerometry, scalp electroencephalography (EEG) and iEEG monitoring. The video, scalp EEG, and accelerometry recordings were manually scored by a board-certified sleep expert into sleep-wake state categories: awake, rapid-eye-movement (REM) sleep, and three non-REM sleep categories (NREM1, 2, 3). The expert labels were used to train, validate, and test a fully automated iEEG sleep-wake classifier in freely behaving canines.Main results. The iEEG-based classifier achieved an overall classification accuracy of 0.878 ± 0.055 and a Cohen's Kappa score of 0.786 ± 0.090. Subsequently, we used the automated iEEG-based classifier to investigate sleep over multiple weeks in freely behaving canines. The results show that the dogs spend a significant amount of the day sleeping, but the characteristics of daytime nap sleep differ from night-time sleep in three key characteristics: during the day, there are fewer NREM sleep cycles (10.81 ± 2.34 cycles per day vs. 22.39 ± 3.88 cycles per night;p< 0.001), shorter NREM cycle durations (13.83 ± 8.50 min per day vs. 15.09 ± 8.55 min per night;p< 0.001), and dogs spend a greater proportion of sleep time in NREM sleep and less time in REM sleep compared to night-time sleep (NREM 0.88 ± 0.09, REM 0.12 ± 0.09 per day vs. NREM 0.80 ± 0.08, REM 0.20 ± 0.08 per night;p< 0.001).Significance.These results support the feasibility and accuracy of automated iEEG sleep-wake classifiers for canine behavior investigations.

New bands in the sleep stages of spider monkeys (Ateles geoffroyi): Electroencephalographic correlations and spatial distribution.

The study of electroencephalographic (EEG) signals in nonhuman primates has led to important discoveries in neurophysiology and sleep behavior. Several studies have analyzed digital EEG data from primate species with prehensile tails, like the spider monkey, and principal component analysis has led to the identification of new EEG bands and their spatial distribution during sleep and wakefulness in these monkeys. However, the spatial location of the EEG correlations of these new bands during the sleep-wake cycle in the spider monkey has not yet been explored. Thus, the objective of this study was to determine the spatial distribution of EEG correlations in the new bands during wakefulness, rapid eye movement (REM) sleep, and non-REM sleep in this species. EEG signals were obtained from the scalp of six monkeys housed in experimental conditions in a laboratory setting. Regarding the 1-21 Hz band, a significant correlation between left frontal and central regions was recorded during non-REM 2 sleep. In the REM sleep, a significant correlation between these cortical areas was seen in two bands: 1-3 and 3-13 Hz. This reflects a modification of the degree of coupling between the cortical areas studied, associated with the distinct stages of sleep. The intrahemispheric EEG correlation found between left perceptual and motor regions during sleep in the spider monkey could indicate activation of a neural circuit for the processing of environmental information that plays a critical role in monitoring the danger of nocturnal predation.

The function of REM and NREM sleep on memory distortion and consolidation.

During rapid eye movement (REM) sleep, newly consolidated memories can be distorted to adjust the existing memory base in memory integration. However, only a few studies have demonstrated the role of REM sleep in memory distortion. The present study aims to clarify the role of REM sleep in the facilitation of memory distortion, that is, hindsight bias, compared to non-rapid eye movement (NREM) sleep and wake states. The split-night paradigm was used to segregate REM and NREM sleep. The hypotheses are (1) hindsight bias-memory distortion-is more substantial during REM-rich sleep (late-night sleep) than during NREM-rich sleep (early-night sleep); (2) memory stabilization is more substantial during NREM-rich sleep (early-night sleep) than during REM-rich sleep (late-night sleep); and (3) memory distortion takes longer time than memory stabilization. The results of the hindsight bias test show that more memory distortions were observed after the REM condition in comparison to the NREM condition. Contrary to the hindsight bias, the correct response in the word-pair association test was observed more in the NREM than in the REM condition. The difference in the hindsight bias index between the REM and NREM conditions was identified only one week later. Comparatively, the difference in correct responses in the word-pair association task between the conditions appeared three hours later and one week later. The present study found that (1) memory distortion occurs more during REM-rich sleep than during NREM-rich sleep, while memory stabilization occurs more during NREM-rich sleep than during REM-rich sleep. Moreover, (2) the newly encoded memory could be stabilized immediately after encoding, but memory distortion occurs over several days. These results suggest that the roles of NREM and REM sleep in memory processes could be different.

A system based on machine learning for improving sleep.

Closed-loop auditory stimulation is one of the well-known and emerging sensory stimulation techniques, which achieves the purpose of sleep regulation by driving the EEG slow oscillation (SO, <1 Hz) through auditory stimulation. The main challenge is to accurately identify the stimulation timing and provide feedback in real-time, which has high requirements on the response time and recognition accuracy of the closed-loop auditory stimulation system. To reduce the impact of systematic errors on the regulation results, most traditional closed-loop auditory stimulation systems try to identify a single feature to determine the timing of stimulus delivery and reduce the system feedback delay by simplifying the calculation. Unlike existing closed-loop regulation systems that identify specific brain features, this paper proposes a closed-loop auditory stimulation sleep regulation system deploying machine learning. The process is: through online sleep real-time automatic staging, tracking the sleep stage to provide feedback quickly, and continuously offering external auditory stimulation at a specific SO phase. This paper uses this system to conduct sleep auditory stimulation regulation experiments on ten subjects. The experimental results show that the sleep closed-loop regulation system proposed in this paper can achieve consistency (effective for almost all subjects in the experiment) and immediate (taking effect immediately after stimulation) modulation effects on SOs. More importantly, the proposed method is superior to existing advanced methods. Therefore, the system designed in this paper has great potential to be more reliable and flexible in sleep regulation.

Automatic sleep-stage classification of heart rate and actigraphy data using deep and transfer learning approaches.

Manual sleep-stage scoring based on full-night polysomnography data recorded in a sleep lab has been the gold standard of clinical sleep medicine. This costly and time-consuming approach is unfit for long-term studies as well as assessment of sleep on a population level. With the vast amount of physiological data becoming available from wrist-worn devices, deep learning techniques provide an opportunity for fast and reliable automatic sleep-stage classification tasks. However, training a deep neural network requires large annotated sleep databases, which are not available for long-term epidemiological studies. In this paper, we introduce an end-to-end temporal convolutional neural network able to automatically score sleep stages from raw heartbeat RR interval (RRI) and wrist actigraphy data. Moreover, a transfer learning approach enables the training of the network on a large public database (Sleep Heart Health Study, SHHS) and its subsequent application to a much smaller database recorded by a wristband device. The transfer learning significantly shortens training time and improves sleep-scoring accuracy from 68.9% to 73.8% and inter-rater reliability (Cohen's kappa) from 0.51 to 0.59. We also found that for the SHHS database, automatic sleep-scoring accuracy using deep learning shows a logarithmic relationship with the training size. Although deep learning approaches for automatic sleep scoring are not yet comparable to the inter-rater reliability among sleep technicians, performance is expected to significantly improve in the near future when more large public databases become available. We anticipate those deep learning techniques, when combined with our transfer learning approach, will leverage automatic sleep scoring of physiological data from wearable devices and enable the investigation of sleep in large cohort studies.

The maturation of aperiodic EEG activity across development reveals a progressive differentiation of wakefulness from sleep.

During development, the brain undergoes radical structural and functional changes following a posterior-to-anterior gradient, associated with profound changes of cortical electrical activity during both wakefulness and sleep. However, a systematic assessment of the developmental effects on aperiodic EEG activity maturation across vigilance states is lacking, particularly regarding its topographical aspects. Here, in a population of 160 healthy infants, children and teenagers (from 2 to 17 years, 10 subjects for each year), we investigated the development of aperiodic EEG activity in wakefulness and sleep. Specifically, we parameterized the shape of the aperiodic background of the EEG Power Spectral Density (PSD) by means of the spectral exponent and offset; the exponent reflects the rate of exponential decay of power over increasing frequencies and the offset reflects an estimate of the y-intercept of the PSD. We found that sleep and development caused the EEG-PSD to rotate over opposite directions: during wakefulness the PSD showed a flatter decay and reduced offset over development, while during sleep it showed a steeper decay and a higher offset as sleep becomes deeper. During deep sleep (N2, N3) only the spectral offset decreased over age, indexing a broad-band voltage reduction. As a result, the difference between values in deep sleep and those in both light sleep (N1) and wakefulness increased with age, suggesting a progressive differentiation of wakefulness from sleep EEG activity, most prominent over the frontal regions, the latest to complete maturation. Notably, the broad-band spectral exponent values during deep sleep stages were entirely separated from wakefulness values, consistently across developmental ages and in line with previous findings in adults. Concerning topographical development, the location showing the steepest PSD decay and largest offset shifted from posterior to anterior regions with age. This shift, particularly evident during deep sleep, paralleled the migration of sleep slow wave activity and was consistent with neuroanatomical and cognitive development. Overall, aperiodic EEG activity distinguishes wakefulness from sleep regardless of age; while, during development, it reveals a postero-anterior topographical maturation and a progressive differentiation of wakefulness from sleep. Our study could help to interpret changes due to pathological conditions and may elucidate the neurophysiological processes underlying the development of wakefulness and sleep.

Preliminary evidence of a relationship between sleep spindles and treatment response in epileptic encephalopathy.

OBJECTIVE: Epileptic encephalopathy with spike-wave activation in sleep (EE-SWAS) is a challenging neurodevelopmental disease characterized by abundant epileptiform spikes during non-rapid eye movement (NREM) sleep accompanied by cognitive dysfunction. The mechanism of cognitive dysfunction is unknown, but treatment with high-dose diazepam may improve symptoms. Spike rate does not predict treatment response, but spikes may disrupt sleep spindles. We hypothesized that in patients with EE-SWAS: (1) spikes and spindles would be anti-correlated, (2) high-dose diazepam would increase spindles and decrease spikes, and (3) spindle response would be greater in those with cognitive improvement. METHODS: Consecutive EE-SWAS patients treated with high-dose diazepam that met the criteria were included. Using a validated automated spindle detector, spindle rate, duration, and percentage were computed in pre- and post-treatment NREM sleep. Spikes were quantified using a validated automated spike detector. The cognitive response was determined from a chart review. RESULTS: Spindle rate was anti-correlated with the spike rate in the channel with the maximal spike rate (p = 0.002) and averaged across all channels (p = 0.0005). Spindle rate, duration, and percentage each increased, and spike rate decreased, after high-dose diazepam treatment (p ≤ 2e-5, all tests). Spindle rate, duration, and percentage (p ≤ 0.004, all tests) were increased in patients with cognitive improvement after treatment, but not those without. Changes in spindle rate but not changes in spike rate distinguished between groups. INTERPRETATION: These findings confirm thalamocortical disruption in EE-SWAS, identify a mechanism through which benzodiazepines may support cognitive recovery, and introduce sleep spindles as a promising mechanistic biomarker to detect treatment response in severe epileptic encephalopathies.

Multi-sensor driver monitoring for drowsiness prediction.

OBJECTIVE: Driver monitoring systems are growing in importance as well as capability. This paper reports drowsy driving detection models that use vehicular, behavioral, and physiological data. The objectives were to augment camera-based system with vehicle-based and heart rate variability measures from a wearable device and compare the performance of drowsiness detection models that use these data sources. Timeliness of the models in predicting drowsiness is analyzed. Timeliness refers to how quickly a model can identify drowsiness and, by extension, how far in advance of an adverse event a classification can be given. METHODS: Behavioral data were provided by a production-type Driver Monitoring System manufactured by Aisin Technical Center of America. Vehicular data were recorded from the National Advanced Driving Simulator's large-excursion motion-base driving simulator. Physiological data were collected from an Empatica E4 wristband. Forty participants drove the simulator for up to three hours after being awake for at least 16 hours. Periodic measurements of drowsiness were recorded every ten minutes using both observational rating of drowsiness by an external rater and the self-reported Karolinska Sleepiness Scale. Nine binary random forest models were created, using different combinations of data sources and ground truths. RESULTS: The classification accuracy of the nine models ranged from 0.77 to 0.92 on a scale from 0 to 1, with 1 indicating a perfect model. The best-performing model included physiological data and used a reduced dataset that eliminated missing data segments after heartrate variability measures were computed. The most timely model was able to detect the presence of drowsiness 6.7 minutes before a drowsy lane departure. CONCLUSIONS: The addition of physiological measures added a small amount of accuracy to the model performance. Models trained on observational ratings of drowsiness detected drowsiness earlier than those based only on Karolinska Sleepiness Scale, making them more timely in detecting the onset of drowsiness.

Identification of major depression patients using machine learning models based on heart rate variability during sleep stages for pre-hospital screening.

With the COVID-19 pandemic causing challenges in hospital admissions globally, the role of home health monitoring in aiding the diagnosis of mental health disorders has become increasingly important. This paper proposes an interpretable machine learning solution to optimise initial screening for major depressive disorder (MDD) in both male and female patients. The data is from the Stanford Technical Analysis and Sleep Genome Study (STAGES). We analyzed 5-min short-term electrocardiogram (ECG) signals during nighttime sleep stages of 40 MDD patients and 40 healthy controls, with a 1:1 gender ratio. After preprocessing, we calculated the time-frequency parameters of heart rate variability (HRV) based on the ECG signals and used common machine learning algorithms for classification, along with feature importance analysis for global decision analysis. Ultimately, the Bayesian optimised extremely randomized trees classifier (BO-ERTC) showed the best performance on this dataset (accuracy 86.32%, specificity 86.49%, sensitivity 85.85%, F1-score 0.86). By using feature importance analysis on the cases confirmed by BO-ERTC, we found that gender is one of the most important factors affecting the prediction of the model, which should not be overlooked in our assisted diagnosis. This method can be embedded in portable ECG monitoring systems and is consistent with the literature results.

Automated sleep staging in people with intellectual disabilities using heart rate and respiration variability.

BACKGROUND: People with intellectual disabilities (ID) have a higher risk of sleep disorders. Polysomnography (PSG) remains the diagnostic gold standard in sleep medicine. However, PSG in people with ID can be challenging, as sensors can be burdensome and have a negative influence on sleep. Alternative methods of assessing sleep have been proposed that could potentially transfer to less obtrusive monitoring devices. The goal of this study was to investigate whether analysis of heart rate variability and respiration variability is suitable for the automatic scoring of sleep stages in sleep-disordered people with ID. METHODS: Manually scored sleep stages in PSGs of 73 people with ID (borderline to profound) were compared with the scoring of sleep stages by the CardioRespiratory Sleep Staging (CReSS) algorithm. CReSS uses cardiac and/or respiratory input to score the different sleep stages. Performance of the algorithm was analysed using input from electrocardiogram (ECG), respiratory effort and a combination of both. Agreement was determined by means of epoch-per-epoch Cohen's kappa coefficient. The influence of demographics, comorbidities and potential manual scoring difficulties (based on comments in the PSG report) was explored. RESULTS: The use of CReSS with combination of both ECG and respiratory effort provided the best agreement in scoring sleep and wake when compared with manually scored PSG (PSG versus ECG = kappa 0.56, PSG versus respiratory effort = kappa 0.53 and PSG versus both = kappa 0.62). Presence of epilepsy or difficulties in manually scoring sleep stages negatively influenced agreement significantly, but nevertheless, performance remained acceptable. In people with ID without epilepsy, the average kappa approximated that of the general population with sleep disorders. CONCLUSIONS: Using analysis of heart rate and respiration variability, sleep stages can be estimated in people with ID. This could in the future lead to less obtrusive measurements of sleep using, for example, wearables, more suitable to this population.

A computationally efficient algorithm for wearable sleep staging in clinical populations.

This study describes a computationally efficient algorithm for 4-class sleep staging based on cardiac activity and body movements. Using an accelerometer to calculate gross body movements and a reflective photoplethysmographic (PPG) sensor to determine interbeat intervals and a corresponding instantaneous heart rate signal, a neural network was trained to classify between wake, combined N1 and N2, N3 and REM sleep in epochs of 30 s. The classifier was validated on a hold-out set by comparing the output against manually scored sleep stages based on polysomnography (PSG). In addition, the execution time was compared with that of a previously developed heart rate variability (HRV) feature-based sleep staging algorithm. With a median epoch-per-epoch κ of 0.638 and accuracy of 77.8% the algorithm achieved an equivalent performance when compared to the previously developed HRV-based approach, but with a 50-times faster execution time. This shows how a neural network, without leveraging any a priori knowledge of the domain, can automatically "discover" a suitable mapping between cardiac activity and body movements, and sleep stages, even in patients with different sleep pathologies. In addition to the high performance, the reduced complexity of the algorithm makes practical implementation feasible, opening up new avenues in sleep diagnostics.

Regional variability in intracerebral properties of NREM to REM sleep transitions in humans.

Transitions between wake and sleep states show a progressive pattern underpinned by local sleep regulation. In contrast, little evidence is available on non-rapid eye movement (NREM) to rapid eye movement (REM) sleep boundaries, considered as mainly reflecting subcortical regulation. Using polysomnography (PSG) combined with stereoelectroencephalography (SEEG) in humans undergoing epilepsy presurgical evaluation, we explored the dynamics of NREM-to-REM transitions. PSG was used to visually score transitions and identify REM sleep features. SEEG-based local transitions were determined automatically with a machine learning algorithm using features validated for automatic intra-cranial sleep scoring (10.5281/zenodo.7410501). We analyzed 2988 channel-transitions from 29 patients. The average transition time from all intracerebral channels to the first visually marked REM sleep epoch was 8 s ± 1 min 58 s, with a great heterogeneity between brain areas. Transitions were observed first in the lateral occipital cortex, preceding scalp transition by 1 min 57 s ± 2 min 14 s (d = -0.83), and close to the first sawtooth wave marker. Regions with late transitions were the inferior frontal and orbital gyri (1 min 1 s ± 2 min 1 s, d = 0.43, and 1 min 1 s ± 2 min 5 s, d = 0.43, after scalp transition). Intracranial transitions were earlier than scalp transitions as the night advanced (last sleep cycle, d = -0.81). We show a reproducible gradual pattern of REM sleep initiation, suggesting the involvement of cortical mechanisms of regulation. This provides clues for understanding oneiric experiences occurring at the NREM/REM boundary.

A reliable automatic algorithm to score fragmentary myoclonus.

Excessive fragmentary myoclonus (EFM) is an incidental polysomnographic finding requiring documentation of ≥20 minutes of NREM sleep with ≥5 fragmentary myoclonus (FM) potentials per minute. Manual FM scoring is time-consuming and prone to inter-rater variability. This work aimed to validate an automatic algorithm to score FM in whole-night recordings. One expert scorer manually scored FM in the anterior tibialis muscles in 10 polysomnographies of as many subjects. The algorithm consisted of two steps. First, parameters of the automatic leg movement identification algorithm of the BrainRT software (OSG, Belgium) were modified to identify FM-like activity. Second, a post-processing algorithm was implemented to remove FM activity not meeting sufficient amplitude criteria. The parameter choice and the post-processing were optimised with leave-one-out cross-validation. Agreement with the human scorer was measured with Cohen's kappa (k) and correlation between manual and automatic FM indices in different sleep stages was evaluated. Agreement in identifying patients with EFM was computed. The algorithm showed substantial agreement (average k > 0.62) for all sleep stages, except for W, where a moderate agreement was observed (average k = 0.58). Nonetheless, the agreement between human scorer and the algorithm was similar to previously reported values of inter-rater variability for FM scoring. Correlation coefficients were over 0.96 for all sleep stages. Furthermore, the presence/absence of EFM was correctly identified in 80% of the subjects. In conclusion, this work presents a reliable algorithm for automatic scoring of FM and EFM. Future studies will apply it to objectively and consistently evaluate FM indices and the presence of EFM in large populations.

Convergent validity of video-based observer rating of drowsiness, against subjective, behavioral, and physiological measures.

Driver drowsiness is a widely recognized cause of motor vehicle accidents. Therefore, a reduction in drowsy driving crashes is required. Many studies evaluating the crash risk of drowsy driving and developing drowsiness detection systems, have used observer rating of drowsiness (ORD) as a reference standard (i.e. ground truth) of drowsiness. ORD is a method of human raters evaluating the levels of driver drowsiness, by visually observing a driver. Despite the widespread use of ORD, concerns remain regarding its convergent validity, which is supported by the relationship between ORD and other drowsiness measures. The objective of the present study was to validate video-based ORD, by examining correlations between ORD levels and other drowsiness measures. Seventeen participants performed eight sessions of a simulated driving task, verbally responding to Karolinska sleepiness scale (KSS), while infra-red face video, lateral position of the participant's car, eye closure, electrooculography (EOG), and electroencephalography (EEG) were recorded. Three experienced raters evaluated the ORD levels by observing facial videos. The results showed significant positive correlations between the ORD levels and all other drowsiness measures (i.e., KSS, standard deviation of the lateral position of the car, percentage of time occupied by slow eye movement calculated from EOG, EEG alpha power, and EEG theta power). The results support the convergent validity of video-based ORD as a measure of driver drowsiness. This suggests that ORD might be suitable as a ground truth for drowsiness.

Evaluation of consensus sleep stage scoring of dysregulated sleep in Parkinson's disease.

OBJECTIVE: Sleep dysregulation in Parkinson's disease (PD) has been hypothesized to occur, in part, from dysfunction in the basal ganglia-cortical circuit. Assessment of this relationship requires accurate sleep stage determination, a known challenge in this clinical population. Our objective was to optimize the consensus on the sleep staging process and reduce interrater variability in a cohort of advanced PD subjects. METHODS: Fifteen PD subjects were enrolled from three sites in a clinical trial that involved recordings from subthalamic nucleus (STN) deep brain stimulation (DBS) leads (NCT04620551). Video polysomnography (vPSG) data for a total of 45 nights were analyzed. Four experienced scorers independently scored data on initial review. Epochs with less than 75% consensus were flagged for secondary review. In secondary review of discordant epochs, two of the original scorers re-assessed epochs, from which the final consensus stage was derived. RESULTS: Sleep stage classification agreement averaged 83.10% across all sleep stages on initial scoring (IS), and on secondary consensus scoring (CS) review, agreement reached 96.58%. Greatest disagreement was noted in determination of awake epochs (33.6% of discordant epochs) and non-rapid-eye-movement stage 2 (N2) epochs (31.8% of discordant epochs). Scoring discrepancy was resolved with direct measurement of cortical frequency and amplitudes, physiologic context of the epoch, and video review. CONCLUSION: Our method of multi-level initial and then secondary consensus review scoring resulted in consensus scoring agreement superior to conventional standards. This work features a custom-engineered vPSG software and review platform for integration of consensus sleep stage scoring in a multi-site clinical trial.

Neural Cross-Frequency Coupling Functions in Sleep.

The human brain presents a heavily connected complex system. From a relatively fixed anatomy, it can enable a vast repertoire of functions. One important brain function is the process of natural sleep, which alters consciousness and voluntary muscle activity. On neural level, these alterations are accompanied by changes of the brain connectivity. In order to reveal the changes of connectivity associated with sleep, we present a methodological framework for reconstruction and assessment of functional interaction mechanisms. By analyzing EEG (electroencephalogram) recordings from human whole night sleep, first, we applied a time-frequency wavelet transform to study the existence and strength of brainwave oscillations. Then we applied a dynamical Bayesian inference on the phase dynamics in the presence of noise. With this method we reconstructed the cross-frequency coupling functions, which revealed the mechanism of how the interactions occur and manifest. We focus our analysis on the delta-alpha coupling function and observe how this cross-frequency coupling changes during the different sleep stages. The results demonstrated that the delta-alpha coupling function was increasing gradually from Awake to NREM3 (non-rapid eye movement), but only during NREM2 and NREM3 deep sleep it was significant in respect of surrogate data testing. The analysis on the spatially distributed connections showed that this significance is strong only for within the single electrode region and in the front-to-back direction. The presented methodological framework is for the whole-night sleep recordings, but it also carries general implications for other global neural states.

AI-Driven sleep staging from actigraphy and heart rate.

Sleep is an important indicator of a person's health, and its accurate and cost-effective quantification is of great value in healthcare. The gold standard for sleep assessment and the clinical diagnosis of sleep disorders is polysomnography (PSG). However, PSG requires an overnight clinic visit and trained technicians to score the obtained multimodality data. Wrist-worn consumer devices, such as smartwatches, are a promising alternative to PSG because of their small form factor, continuous monitoring capability, and popularity. Unlike PSG, however, wearables-derived data are noisier and far less information-rich because of the fewer number of modalities and less accurate measurements due to their small form factor. Given these challenges, most consumer devices perform two-stage (i.e., sleep-wake) classification, which is inadequate for deep insights into a person's sleep health. The challenging multi-class (three, four, or five-class) staging of sleep using data from wrist-worn wearables remains unresolved. The difference in the data quality between consumer-grade wearables and lab-grade clinical equipment is the motivation behind this study. In this paper, we present an artificial intelligence (AI) technique termed sequence-to-sequence LSTM for automated mobile sleep staging (SLAMSS), which can perform three-class (wake, NREM, REM) and four-class (wake, light, deep, REM) sleep classification from activity (i.e., wrist-accelerometry-derived locomotion) and two coarse heart rate measures-both of which can be reliably obtained from a consumer-grade wrist-wearable device. Our method relies on raw time-series datasets and obviates the need for manual feature selection. We validated our model using actigraphy and coarse heart rate data from two independent study populations: the Multi-Ethnic Study of Atherosclerosis (MESA; N = 808) cohort and the Osteoporotic Fractures in Men (MrOS; N = 817) cohort. SLAMSS achieves an overall accuracy of 79%, weighted F1 score of 0.80, 77% sensitivity, and 89% specificity for three-class sleep staging and an overall accuracy of 70-72%, weighted F1 score of 0.72-0.73, 64-66% sensitivity, and 89-90% specificity for four-class sleep staging in the MESA cohort. It yielded an overall accuracy of 77%, weighted F1 score of 0.77, 74% sensitivity, and 88% specificity for three-class sleep staging and an overall accuracy of 68-69%, weighted F1 score of 0.68-0.69, 60-63% sensitivity, and 88-89% specificity for four-class sleep staging in the MrOS cohort. These results were achieved with feature-poor inputs with a low temporal resolution. In addition, we extended our three-class staging model to an unrelated Apple Watch dataset. Importantly, SLAMSS predicts the duration of each sleep stage with high accuracy. This is especially significant for four-class sleep staging, where deep sleep is severely underrepresented. We show that, by appropriately choosing the loss function to address the inherent class imbalance, our method can accurately estimate deep sleep time (SLAMSS/MESA: 0.61±0.69 hours, PSG/MESA ground truth: 0.60±0.60 hours; SLAMSS/MrOS: 0.53±0.66 hours, PSG/MrOS ground truth: 0.55±0.57 hours;). Deep sleep quality and quantity are vital metrics and early indicators for a number of diseases. Our method, which enables accurate deep sleep estimation from wearables-derived data, is therefore promising for a variety of clinical applications requiring long-term deep sleep monitoring.

Tfap2b acts in GABAergic neurons to control sleep in mice.

Sleep is a universal state of behavioral quiescence in both vertebrates and invertebrates that is controlled by conserved genes. We previously found that AP2 transcription factors control sleep in C. elegans, Drosophila, and mice. Heterozygous deletion of Tfap2b, one of the mammalian AP2 paralogs, reduces sleep in mice. The cell types and mechanisms through which Tfap2b controls sleep in mammals are, however, not known. In mice, Tfap2b acts during early embryonic stages. In this study, we used RNA-seq to measure the gene expression changes in brains of Tfap2b-/- embryos. Our results indicated that genes related to brain development and patterning were differentially regulated. As many sleep-promoting neurons are known to be GABAergic, we measured the expression of GAD1, GAD2 and Vgat genes in different brain areas of adult Tfap2b+/- mice using qPCR. These experiments suggested that GABAergic genes are downregulated in the cortex, brainstem and cerebellum areas, but upregulated in the striatum. To investigate whether Tfap2b controls sleep through GABAergic neurons, we specifically deleted Tfap2b in GABAergic neurons. We recorded the EEG and EMG before and after a 6-h period of sleep deprivation and extracted the time spent in NREM and in REM sleep as well as delta and theta power to assess NREM and REM sleep, respectively. During baseline conditions, Vgat-tfap2b-/- mice exhibited both shortened NREM and REM sleep time and reduced delta and theta power. Consistently, weaker delta and theta power were observed during rebound sleep in the Vgat-tfap2b-/- mice after sleep deprivation. Taken together, the results indicate that Tfap2b in GABAergic neurons is required for normal sleep.

A novel machine learning system for identifying sleep-wake states in mice.

Research into sleep-wake behaviors relies on scoring sleep states, normally done by manual inspection of electroencephalogram (EEG) and electromyogram (EMG) recordings. This is a highly time-consuming process prone to inter-rater variability. When studying relationships between sleep and motor function, analyzing arousal states under a four-state system of active wake (AW), quiet wake (QW), nonrapid-eye-movement (NREM) sleep, and rapid-eye-movement (REM) sleep provides greater precision in behavioral analysis but is a more complex model for classification than the traditional three-state identification (wake, NREM, and REM sleep) usually used in rodent models. Characteristic features between sleep-wake states provide potential for the use of machine learning to automate classification. Here, we devised SleepEns, which uses a novel ensemble architecture, the time-series ensemble. SleepEns achieved 90% accuracy to the source expert, which was statistically similar to the performance of two other human experts. Considering the capacity for classification disagreements that are still physiologically reasonable, SleepEns had an acceptable performance of 99% accuracy, as determined blindly by the source expert. Classifications given by SleepEns also maintained similar sleep-wake characteristics compared to expert classifications, some of which were essential for sleep-wake identification. Hence, our approach achieves results comparable to human ability in a fraction of the time. This new machine-learning ensemble will significantly impact the ability of sleep researcher to detect and study sleep-wake behaviors in mice and potentially in humans.