Conventional Polysomnography Research Articles

Deep learning approaches for assessing pediatric sleep apnea severity through SpO2 signals

Pediatric Sleep Apnea–Hypopnea (SAH) presents a significant health challenge, particularly in diagnostic contexts, where conventional Polysomnography (PSG) testing, although effective, can be distressing for children. Addressing this, our research proposes a less invasive method to assess pediatric SAH severity by analyzing blood oxygen saturation (SpO2) signals. We adopted two advanced deep learning architectures, namely ResNet-based and attention-augmented hybrid CNN-BiGRU models, to process SpO2 signals in a one-dimensional (1D) format for Apnea–Hypopnea Index (AHI) estimation in pediatric subjects. Employing the CHAT dataset, which includes 844 SpO2 signals, the data was partitioned into training (60%), testing (30%), and validation (10%) sets. A predefined validation subset was randomly selected to ensure the models' robustness via a threefold cross-validation approach. Comparative analysis revealed that while the ResNet model attained an average accuracy of 72.9% across four SAH severity categories with a kappa score of 0.57, the CNN-BiGRU-Attention model demonstrated superior performance, achieving an average accuracy of 75.95% and a kappa score of 0.63. This distinction underscores our method's efficacy in both estimating AHI and categorizing SAH severity levels with notable precision. Further, to evaluate diagnostic capabilities, the models were benchmarked against common AHI thresholds (1, 5, and 10 events/hour) in each test fold, affirming their effectiveness in identifying pediatric SAH. This study marks a significant advance in the field, offering a non-invasive, child-friendly alternative for pediatric SAH diagnosis. Although challenges persist in accurately estimating AHI, particularly in severe cases, our findings represent a critical stride towards improving diagnostic processes in pediatric SAH.

Exploring the potential of a new wearable sleep monitoring device for clinical application

OBJECTIVESSleep is critical to health and quality of life, and the gold-standard assessment method for sleep disorders--polysomnography (PSG)--has struggled to become widely used because of its inherent shortcomings, which portable sleep-monitoring devices are expected to improve. This study aims to validate a new EEG-, EOG-, and EMG-based portable sleep-monitoring device against conventional PSG to explore its feasibility for clinical use. MethodsTwenty participants completed overnight sleep monitoring using conventional polysomnography and the Wearable PolysomnogramTM (WPSG-I). The similarity of the raw signals recorded by the two devices and the agreement rate of their sleep-staging results were compared. ResultsThe raw electroencephalogram (EEG) signals recorded by the two devices were highly similar, and the agreement rates between WPSG-I manual sleep staging and automatic sleep staging with conventional PSG manual sleep staging were 95.8 % (kappa = 0.92) and 89.7 % (kappa = 0.80), respectively. ConclusionThe agreement rate of sleep staging results between the two devices was high. The results of this study demonstrate the validity and feasibility of the WPSG-I for sleep research and clinical applications.

Predicting Alzheimer's disease CSF core biomarkers: a multimodal Machine Learning approach.

Alzheimer's disease (AD) is a progressive neurodegenerative disorder. Current core cerebrospinal fluid (CSF) AD biomarkers, widely employed for diagnosis, require a lumbar puncture to be performed, making them impractical as screening tools. Considering the role of sleep disturbances in AD, recent research suggests quantitative sleep electroencephalography features as potential non-invasive biomarkers of AD pathology. However, quantitative analysis of comprehensive polysomnography (PSG) signals remains relatively understudied. PSG is a non-invasive test enabling qualitative and quantitative analysis of a wide range of parameters, offering additional insights alongside other biomarkers. Machine Learning (ML) gained interest for its ability to discern intricate patterns within complex datasets, offering promise in AD neuropathology detection. Therefore, this study aims to evaluate the effectiveness of a multimodal ML approach in predicting core AD CSF biomarkers. Mild-moderate AD patients were prospectively recruited for PSG, followed by testing of CSF and blood samples for biomarkers. PSG signals underwent preprocessing to extract non-linear, time domain and frequency domain statistics quantitative features. Multiple ML algorithms were trained using four subsets of input features: clinical variables (CLINVAR), conventional PSG parameters (SLEEPVAR), quantitative PSG signal features (PSGVAR) and a combination of all subsets (ALL). Cross-validation techniques were employed to evaluate model performance and ensure generalizability. Regression models were developed to determine the most effective variable combinations for explaining variance in the biomarkers. On 49 subjects, Gradient Boosting Regressors achieved the best results in estimating biomarkers levels, using different loss functions for each biomarker: least absolute deviation (LAD) for the Aβ42, least squares (LS) for p-tau and Huber for t-tau. The ALL subset demonstrated the lowest training errors for all three biomarkers, albeit with varying test performance. Specifically, the SLEEPVAR subset yielded the best test performance in predicting Aβ42, while the ALL subset most accurately predicted p-tau and t-tau due to the lowest test errors. Multimodal ML can help predict the outcome of CSF biomarkers in early AD by utilizing non-invasive and economically feasible variables. The integration of computational models into medical practice offers a promising tool for the screening of patients at risk of AD, potentially guiding clinical decisions.

Deglutition dynamics of patients with obstructive sleep apnea.

In the present study, we performed a detailed analysis of deglutitive dynamics during sleep in patients with obstructive sleep apnea (OSA) using a methodology developed by Sato etal. We hypothesized that the frequency of deglutition would decrease with increasing severity of OSA. The aim of this study is to clarify the involvement of deglutitive dynamics during sleep in OSA by investigating the correlations between deglutition and sleep parameters. This study included 30 adult patients with OSA. To analyze deglutition dynamics during sleep, surface electromyography recordings of the suprahyoid and thyrohyoid neck muscles, which are involved in deglutition, were performed simultaneous with conventional polysomnography. The "index of deglutition" was defined as the frequency of deglutition per hour of sleep. We examined correlations between this index and sleep parameters (apnea-hypopnea index [AHI], apnea index, hypopnea index, and lowest blood oxygen saturation). By analyzing the obtained polysomnography and electromyography waveforms, we identified two deglutition patterns with and without respiratory arousal during sleep. We found a significant negative correlation between the index of deglutition in sleep stage 1 and the AHI, with a correlation coefficient of -0.48. (p=0.02). In the current study, we distinguished deglutition during sleep with and without arousal. In addition we discovered a significant negative correlation between the index of deglutition in sleep stage 1 and the AHI. This new finding will provide a platform for future research on OSA in aspiration pneumonia.

Diagnostic accuracy of daytime polysomnography: a reappraisal during the COVID-19 era.

Level I conventional polysomnography (PSG), the gold standard for diagnosing obstructive sleep apnea (OSA), requires an overnight stay. This study evaluated the role of daytime PSG as an alternative diagnostic tool. A prospective cohort study was undertaken with consecutive patients with suspected OSA at a tertiary care sleep center. The primary objective was to evaluate the sensitivity and diagnostic accuracy of daytime PSG for diagnosing OSA. The secondary objective was to find out the factors associated with a falsely negative daytime PSG result. All individuals were subjected to level I daytime PSG, done in the sleep lab in the presence of an experienced sleep technician during the daytime from 12 PM to 4 PM. Out of 162 patients, 105 underwent daytime PSG. OSA was diagnosed on daytime PSG in 86.7 out of the 19 remaining patients refused a repeat PSG study. Out of the 12 individuals who underwent the nighttime PSG for confirmatory diagnosis, 10 were diagnosed as OSA (false negatives), and 2 were confirmed as not-OSA (true negatives). The sensitivity, diagnostic accuracy, and negative predictive value of daytime PSG were 89.58%, 89.80%, and 16.67%, respectively. The false negatives had a higher prevalence of mild OSA. Daytime PSG is sensitive in diagnosing OSA and can be considered in individuals with severe symptoms at centers with a high patient load or when the individual wishes to avoid a nighttime study. A negative result in daytime PSG must be followed by conventional overnight PSG for confirmatory diagnosis.

Application of LightGBM hybrid model based on TPE algorithm optimization in sleep apnea detection.

Sleep apnoea syndrome (SAS) is a serious sleep disorder and early detection of sleep apnoea not only reduces treatment costs but also saves lives. Conventional polysomnography (PSG) is widely regarded as the gold standard diagnostic tool for sleep apnoea. However, this method is expensive, time-consuming and inherently disruptive to sleep. Recent studies have pointed out that ECG analysis is a simple and effective diagnostic method for sleep apnea, which can effectively provide physicians with an aid to diagnosis and reduce patients' suffering. To this end, in this paper proposes a LightGBM hybrid model based on ECG signals for efficient detection of sleep apnea. Firstly, the improved Isolated Forest algorithm is introduced to remove abnormal data and solve the data sample imbalance problem. Secondly, the parameters of LightGBM algorithm are optimised by the improved TPE (Tree-structured Parzen Estimator) algorithm to determine the best parameter configuration of the model. Finally, the fusion model TPE_OptGBM is used to detect sleep apnoea. In the experimental phase, we validated the model based on the sleep apnoea ECG database provided by Phillips-University of Marburg, Germany. The experimental results show that the model proposed in this paper achieves an accuracy of 95.08%, a precision of 94.80%, a recall of 97.51%, and an F1 value of 96.14%. All of these evaluation indicators are better than the current mainstream models, which is expected to assist the doctor's diagnostic process and provide a better medical experience for patients.

O016 Caregivers’ Perspectives on the Sonomat as a Polysomnography Alternative in Children with Down syndrome

Abstract Introduction Children with Down syndrome (DS) frequently require repeated polysomnography (PSG) to diagnose and monitor sleep disordered breathing (SDB). PSG set-up can be distressing for this population, who may struggle with sensor and lead application. The Sonomat, a non-invasive mattress overlay, has now been shown to accurately detect and profile SDB in this population. The aim of this study was to evaluate caregiver experiences of undertaking home Sonomat monitoring in their child with DS, compared to conventional lab PSG. Method Children with DS aged 2-17yrs undergoing a diagnostic PSG at the Queensland Children’s Hospital completed a PSG questionnaire about their child’s experience of this test. Those that also undertook home Sonomat monitoring completed a questionnaire about their experience of using the Sonomat in the home. Results Caregivers of fifty-one children with DS (median age 7yrs, 52% male) completed the PSG questionnaire and twenty-five completed the Sonomat questionnaire. 45% (n=51) of caregivers reported that in-lab PSG was ‘hard’ or ‘very hard’ for their child. 100% (n=25) of the caregivers rated their experience of using the Sonomat at home as good, very good or excellent. 88% (n=25) of caregivers felt that the Sonomat was their preferred option for monitoring in children with DS, as opposed to a traditional in-lab PSG. Conclusion In this study, the Sonomat was viewed favourably by caregivers of children with DS, and in the majority of instances was preferred to in-hospital PSG.

L-SeqSleepNet: Whole-cycle Long Sequence Modelling for Automatic Sleep Staging.

Human sleep is cyclical with a period of approximately 90 minutes, implying long temporal dependency in the sleep data. Yet, exploring this long-term dependency when developing sleep staging models has remained untouched. In this work, we show that while encoding the logic of a whole sleep cycle is crucial to improve sleep staging performance, the sequential modelling approach in existing state-of-the-art deep learning models are inefficient for that purpose. We thus introduce a method for efficient long sequence modelling and propose a new deep learning model, L-SeqSleepNet, which takes into account whole-cycle sleep information for sleep staging. Evaluating L-SeqSleepNet on four distinct databases of various sizes, we demonstrate state-of-the-art performance obtained by the model over three different EEG setups, including scalp EEG in conventional Polysomnography (PSG), in-ear EEG, and around-the-ear EEG (cEEGrid), even with a single EEG channel input. Our analyses also show that L-SeqSleepNet is able to alleviate the predominance of N2 sleep (the major class in terms of classification) to bring down errors in other sleep stages. Moreover the network becomes much more robust, meaning that for all subjects where the baseline method had exceptionally poor performance, their performance are improved significantly. Finally, the computation time only grows at a sub-linear rate when the sequence length increases.

Self-applied somnography: technical feasibility of electroencephalography and electro-oculography signal characteristics in sleep staging of suspected sleep-disordered adults.

Sleep recordings are increasingly being conducted in patients' homes where patients apply the sensors themselves according to instructions. However, certain sensor types such as cup electrodes used in conventional polysomnography are unfeasible for self-application. To overcome this, self-applied forehead montages with electroencephalography and electro-oculography sensors have been developed. We evaluated the technical feasibility of a self-applied electrode set from Nox Medical (Reykjavik, Iceland) through home sleep recordings of healthy and suspected sleep-disordered adults (n = 174) in the context of sleep staging. Subjects slept with a double setup of conventional type II polysomnography sensors and self-applied forehead sensors. We found that the self-applied electroencephalography and electro-oculography electrodes had acceptable impedance levels but were more prone to losing proper skin-electrode contact than the conventional cup electrodes. Moreover, the forehead electroencephalography signals recorded using the self-applied electrodes expressed lower amplitudes (difference 25.3%-43.9%, p < 0.001) and less absolute power (at 1-40 Hz, p < 0.001) than the polysomnography electroencephalography signals in all sleep stages. However, the signals recorded with the self-applied electroencephalography electrodes expressed more relative power (p < 0.001) at very low frequencies (0.3-1.0 Hz) in all sleep stages. The electro-oculography signals recorded with the self-applied electrodes expressed comparable characteristics with standard electro-oculography. In conclusion, the results support the technical feasibility of the self-applied electroencephalography and electro-oculography for sleep staging in home sleep recordings, after adjustment for amplitude differences, especially for scoring Stage N3 sleep.

The REM microarousal and REM duration as the potential indicator in paradoxical insomnia

ObjectiveAlthough paradoxical insomnia is a prevalent subtype of chronic insomnia, the etiology of it is unclear. Contrary to complaints of little or no sleep, polysomnography (PSG) findings show that paradoxical insomnia patients have near normal sleep macrostructure. The purpose of this study is to determine the changes of microstructure and explore the etiology of paradoxical insomnia. MethodsThe PSG findings of 89 paradoxical insomnia patients were compared with those of 41 gender balanced healthy controls without sleep complaints. All subjects underwent nocturnal PSG recordings. Conventional PSG measures and microarousals were quantified and statistically analyzed. Receiver operating characteristic curve and correlation analysis were used to evaluate the potential of REM sleep microarousals and REM duration as indicators of paradoxical insomnia. ResultsCompared with the controls, paradoxical insomnia patients had no significant differences in sleep macrostructures. Statistical analysis showed that non-rapid eye movement (NREM) microarousals revealed no significant differences between paradoxical insomnia patients and controls. Noticeably, more spontaneous microarousals appeared in rapid eye movement (REM) stage for paradoxical insomnia patients. Based on receiver operating characteristic curve (ROC), the optimal cutoff value of REM sleep microarousals could predict paradoxical insomnia. Furthermore, a positive correlation between microarousals in REM sleep and the duration of REM sleep was presented in paradoxical insomnia patients. ConclusionsThe frequency of REM microarousals and the duration of REM sleep could reflect the real sleep state of paradoxical insomnia patients. That suggested PSG investigation extended to microarousal could be helpful to understand the etiology in paradoxical insomnia.

How does obstructive sleep apnea alter cerebral hemodynamics?

We aimed to characterize the cerebral hemodynamic response to obstructive sleep apnea/hypopnea events, and evaluate their association to polysomnographic parameters. The characterization of the cerebral hemodynamics in obstructive sleep apnea (OSA) may add complementary information to further the understanding of the severity of the syndrome beyond the conventional polysomnography. Severe OSA patients were studied during night sleep while monitored by polysomnography. Transcranial, bed-side diffuse correlation spectroscopy (DCS) and frequency-domain near-infrared diffuse correlation spectroscopy (NIRS-DOS) were used to follow microvascular cerebral hemodynamics in the frontal lobes of the cerebral cortex. Changes in cerebral blood flow (CBF), total hemoglobin concentration (THC), and cerebral blood oxygen saturation (StO2) were analyzed. We considered 3283 obstructive apnea/hypopnea events from sixteen OSA patients (Age (median, interquartile range) 57 (52-64.5); females 25%; AHI (apnea-hypopnea index) 84.4 (76.1-93.7)). A biphasic response (maximum/minimum followed by a minimum/maximum) was observed for each cerebral hemodynamic variable (CBF, THC, StO2), heart rate and peripheral arterial oxygen saturation (SpO2). Changes of the StO2 followed the dynamics of the SpO2, and were out of phase from the THC and CBF. Longer events were associated with larger CBF changes, faster responses and slower recoveries. Moreover, the extrema of the response to obstructive hypopneas were lower compared to apneas (p < .001). Obstructive apneas/hypopneas cause profound, periodic changes in cerebral hemodynamics, including periods of hyper- and hypo-perfusion and intermittent cerebral hypoxia. The duration of the events is a strong determinant of the cerebral hemodynamic response, which is more pronounced in apnea than hypopnea events.

0503 Respiratory Effort During Sleep And Prevalent Diabetes In Obstructive Sleep Apnea

Abstract Introduction Obstructive sleep apnea (OSA) and type 2 diabetes (T2Dia) is a frequent clinical association. Beyond OSA-related hypoxic burden, increased respiratory effort (RE) is one of the main features of OSA and contributes to sympathetic overactivity that in turn might participate to glucose homeostasis dysregulation. The independent contribution of RE to the OSA-T2Dia pathogenesis remains to be demonstrated. The study aims to determine the impact of metrics assessing sleep RE derived from the sleep mandibular jaw movements (MJM) signal on the prevalence of T2Dia in a large cohort of patients addressed for in-laboratory conventional polysomnography (PSG) with suspicion of OSA. Methods An interpretable machine learning model was built to predict T2Dia from clinical data, PSG indices, and MJM-derived parameters (including the time spent with increased RE, expressed in % of TST). The analysis included 1128 subjects, randomly assigned to training (n=853) and validation (n=275) subsets with equal T2Dia prevalence of 11%. Results The risk stratification model based on 19 input features including increased RE showed good performance for predicting prevalent T2Dia (sensitivity=0.81, specificity=0.89). Post-hoc interpretation using the Shapley additive explanation (SHAP) method revealed that increased RE was the most important risk factor of T2Dia after other clinical factors for T2Dia (i.e., sex, BMI and average SpO2), ahead of standard PSG metrics (including the apnea-hypopnea index and oxygen desaturation index). Conclusion These findings suggested that the proportion of sleep time spent with increased RE automatically derived from MJM is a potential new reliable metric to predict prevalent T2Dia in patients with OSA. Support (if any)

Respiratory effort during sleep and prevalent hypertension in obstructive sleep apnoea.

Mechanisms underlying blood pressure changes in obstructive sleep apnoea (OSA) are incompletely understood. Increased respiratory effort is one of the main features of OSA and is associated with sympathetic overactivity, leading to increased vascular wall stiffness and remodelling. This study investigated associations between a new measure of respiratory effort (percentage of total sleep time spent with increased respiratory effort based on measurement of mandibular jaw movements (MJM): REMOV, %TST) and prevalent hypertension in adults referred for evaluation of suspected OSA. A machine learning model was built to predict hypertension from clinical data, conventional polysomnography (PSG) indices and MJM-derived parameters (including REMOV). The model was evaluated in a training subset and a test subset. The analysis included 1127 patients: 901 (80%) in the training subset and 226 (20%) in the test subset. The prevalence of hypertension was 31% and 30%, respectively, in the training and test subsets. A risk stratification model based on 18 input features including REMOV had good accuracy for predicting prevalent hypertension (sensitivity 0.75 and specificity 0.83). Using the Shapley additive explanation method, REMOV was the best predictor of hypertension after clinical risk factors (age, sex, body mass index and neck circumference) and time with oxygen saturation <90%, ahead of standard PSG metrics (including the apnoea-hypopnoea index and oxygen desaturation index). The proportion of sleep time spent with increased respiratory effort automatically derived from MJM was identified as a potential new reliable metric to predict prevalent hypertension in patients with OSA.

Evaluation of bladder function and sleep patterns in children with primary mono-symptomatic nocturnal enuresis by polysomnography combined with ambulatory urodynamic monitoring

Objective: To evaluate the bladder function and sleep pattern in the children with primary mono-symptomatic nocturnal enuresis (PMNE) by the polysomnography (PSG) and ambulatory urodynamic monitoring (AUM). Methods: From October 2019 to October 2021, forty-three patients with PMNE were selected as PMNE group from the First Affiliated Hospital of Zhengzhou University and further subdivided into the severe PMNE group (enuresis>4 times/week) and the non-severe PMNE group (enuresis times 4 times/week) according to the severity. The conventional urodynamics (CUD), AUM, and PSG examinations and bladder diary were completed in the PMNE group. The control group consisted of 23 children with normal PSG findings and without the lower urinary tract symptoms. Results: The severe PMNE group included 9 males and 14 females, aged(12.1±3.2)years, and nocturnal enuresis number per week is 6.7±1.7. The non-severe PMNE group included 9 males and 11 females, aged(12.0±3.4)years, and nocturnal enuresis number per week is 2.3±1.0. The incidences of nocturnal polyuria and the reduction in maximum bladder capacity in the PMNE group was 34.9% and 11.6%, respectively. The incidence and frequency of detrusor overactivity (DO) in the severe PMNE group were significantly higher than those in the non-severe PMNE group [78.3% vs 45.0%, (5.5±1.8) times/h vs (3.4±1.0) times/h, respectively, all P<0.05]. It was found by the PSG that the severe PMNE group had significantly higher cortical arousal index, apnea hypopnea index (AHI), and percentage of N1+N2 phase in total sleep time, compared with the control group[(58.6±9.8)% vs (49.3±9.5)%, (9.4±4.4) times/h vs (3.1±1.5) times/h, (2.7±0.9) times/h vs (0.9±0.7) times/h] (all P<0.05). While the sleep efficiency of the severe PMNE group was substantially lower than that of the non-severe PMNE group [(86.4±4.3)% vs (91.0±3.9)%], the cortical arousal index and AHI were significantly greater than those of the non-severe PMNE group[(9.4±4.4) times/h vs (5.7±3.2) times/h, (2.7±0.9) times/h vs (1.9±0.7) times/h] (all P<0.05). In the PMNE group, there were positive correlations between cortical arousal index and nocturnal DO frequency or AHI (r=0.705, 0.765, P=0.001). Conclusions: Children with PMNE have nocturnal bladder dysfunction and abnormal sleep pattern, and there is a certain correlation between them. PSG and AUM are necessary for the evaluation and treatment of children with PMNE.

An automated heart rate-based algorithm for sleep stage classification: Validation using conventional polysomnography and an innovative wearable electrocardiogram device.

BackgroundThe rapid advancement in wearable solutions to monitor and score sleep staging has enabled monitoring outside of the conventional clinical settings. However, most of the devices and algorithms lack extensive and independent validation, a fundamental step to ensure robustness, stability, and replicability of the results beyond the training and testing phases. These systems are thought not to be feasible and reliable alternatives to the gold standard, polysomnography (PSG).Materials and methodsThis validation study highlights the accuracy and precision of the proposed heart rate (HR)-based deep-learning algorithm for sleep staging. The illustrated solution can perform classification at 2-levels (Wake; Sleep), 3-levels (Wake; NREM; REM) or 4- levels (Wake; Light; Deep; REM) in 30-s epochs. The algorithm was validated using an open-source dataset of PSG recordings (Physionet CinC dataset, n = 994 participants, 994 recordings) and a proprietary dataset of ECG recordings (Z3Pulse, n = 52 participants, 112 recordings) collected with a chest-worn, wireless sensor and simultaneous PSG collection using SOMNOtouch.ResultsWe evaluated the performance of the models in both datasets in terms of Accuracy (A), Cohen’s kappa (K), Sensitivity (SE), Specificity (SP), Positive Predictive Value (PPV), and Negative Predicted Value (NPV). In the CinC dataset, the highest value of accuracy was achieved by the 2-levels model (0.8797), while the 3-levels model obtained the best value of K (0.6025). The 4-levels model obtained the lowest SE (0.3812) and the highest SP (0.9744) for the classification of Deep sleep segments. AHI and biological sex did not affect scoring, while a significant decrease of performance by age was reported across the models. In the Z3Pulse dataset, the highest value of accuracy was achieved by the 2-levels model (0.8812), whereas the 3-levels model obtained the best value of K (0.611). For classification of the sleep states, the lowest SE (0.6163) and the highest SP (0.9606) were obtained for the classification of Deep sleep segment.ConclusionThe results of the validation procedure demonstrated the feasibility of accurate HR-based sleep staging. The combination of the proposed sleep staging algorithm with an inexpensive HR device, provides a cost-effective and non-invasive solution deployable in the home environment and robust across age, sex, and AHI scores.

Cyclic Sternocleidomastoid Contraction: An Unusual Finding in a Conventional Polysomnography

Cyclic Sternocleidomastoid Contraction: An Unusual Finding in a Conventional Polysomnography

Temporal relations between peripheral and central arousals in good and poor sleepers

Good sleepers and patients with insomnia symptoms (poor sleepers) were tracked with two measures of arousal; conventional polysomnography (PSG) for electroencephalogram (EEG) assessed cortical arousals, and a peripheral arterial tonometry device was used for the detection of peripheral nervous system (PNS) arousals associated with vasoconstrictions. The relationship between central (cortical) and peripheral (autonomic) arousals was examined by evaluating their close temporal dynamics. Cortical arousals almost invariably were preceded and followed by peripheral activations, while large peripheral autonomic arousals were followed by cortical arousals only half of the time. The temporal contiguity of these two types of arousals was altered in poor sleepers, and poor sleepers displayed a higher number of cortical and peripheral arousals compared with good sleepers. Given the difference in the number of peripheral autonomic arousals between good and poor sleepers, an evaluation of such arousals could become a means of physiologically distinguishing poor sleepers.

O038 Lower mean oxygen saturation in sleep is associated with worse cognitive performance in subjects with obstructive sleep apnoea

Abstract Introduction Obstructive sleep apnoea (OSA) is a risk factor for cognitive impairment and has been associated with deficits in executive function, attention, and memory. Potential mechanisms of harm include sleep disruption and intermittent hypoxaemia. Our aim was to investigate whether the apnoea-hypopnoea index (AHI), arousal index (AI) and mean oxyhemoglobin saturation in sleep (mean SpO2) - conventional polysomnography (PSG) measures of respiratory disturbance, sleep fragmentation and nocturnal hypoxaemia respectively - were associated with worse cognitive performance in OSA subjects. Methods In this cross-sectional analysis, 75 subjects with PSG-confirmed OSA (age: 66.1yrs ± 7.1yrs, male: 51%) were recruited from a hospital sleep clinic and had their cognitive profile screened via the Addenbrooke’s Cognitive Examination – Revised (ACE-R). Linear regression was used to determine whether AHI, AI and mean SpO2 were associated with total ACE-R scores. Binary logistic regressions were then performed to determine whether increased severity of OSA (AHI ≥ 30 events/hour), sleep fragmentation (AI ≥ 30 events/hour), and hypoxaemia (mean SpO2 ≤ 92%) increased the likelihood that participants would have worse cognition (ACE-R score ≤ 88). Results There was a modest positive association with mean SpO2 and ACE-R score (r² = 10.4%, p &amp;lt; 0.01). Similarly, logistic regression found only increased hypoxaemia (mean SpO2 ≤ 92%) to be associated with increased odds of worsened cognition (OR 3.00, 95% CI (1.090–8.254), p &amp;lt; 0.05). Discussion OSA-induced hypoxaemia, and not sleep fragmentation or respiratory disturbance, was found to be most strongly associated with deficits in cognitive performance.

666 Sleep Architecture in the Intensive Care Unit As Revealed via Breathing and Heart Rate Variability

Abstract Introduction Sleep in the intensive care unit (ICU) is difficult to measure by conventional polysomnography. We investigated the feasibility of assessing sleep state from readily available ICU signals: heart rate variability (HRV) from electrocardiography and breathing from a wearable respiratory band. We compared findings with an age and sex matched sleep laboratory group. Methods As part of a clinical trial, 102 adult non-ventilated patients in three ICUs in the Massachusetts General Hospital wore a respiratory band. Both heart rate variability (RR-intervals) from ECG, and breathing (respiratory effort waveforms) data for up to seven days per patient were obtained. 220 age- and sex-matched subjects from a sleep lab cohort who wore the same respiratory effort band and ECG were selected for comparison. We staged sleep from the HRV and breathing data using previously published deep neural network models. We defined discordant sleep epochs as those where HRV- and breathing-based models disagreed. Agreement was computed for the following pairs: (R,R),(N1,N1),(N2,N2),(N3,N3),(N1,W),(N1,N2),(N2,N3). Results Demographics: Mean(STD) age: ICU 68(9), sleeplab 68(9); BMI: ICU 27(6), sleeplab 31(6); ICU 40% female, sleeplab 44% female; race: ICU%:Sleeplab% 90:69 White, 5:4 Black, 2:7 Asian. 34% of ICU-subjects were in a medical ICU, 66% in a surgical ICU. Mean total sleep duration in the ICU was 8.9 hours (4.5h concordant, 4.4h discordant sleep). We observed increased amounts of discordant sleep in the ICU compared with the sleeplab cohort (4.4h vs. 1h, p&amp;lt;0.01). We found different REM sleep distributions (p&amp;lt;0.01) with reduced median (10% vs. 20%) but elevated 90% quantile (45% vs. 26%), elevated N1(%) (41% vs. 26%, p&amp;lt;0.05), reduced N2(%) (19 vs. 44, p&amp;lt;0.01), and reduced N2+N3(%) (34 vs. 59, p&amp;lt;0.05). We further observed higher mean respiratory rate (17.4 vs. 15.9 breaths per minute, p&amp;lt;0.01), lower inter-breath-intervals (3.9 vs. 4.7 seconds per breath, p&amp;lt;0.01), and more breathing variability than in sleeplab AHI&amp;lt;5 group but less than in AHI&amp;gt;15 group. Conclusion HRV and respiratory-based measures can assess sleep in the ICU. The findings of increased discordant sleep in the ICU might stem from limitations of the models, fundamental changes in sleep biology during critical illness, pharmaceutical drugs, sleep fragmentation, and/or associated pathology in the ICU. Support (if any):

391 Upper Airway Resistance Syndrome is associated with high Cyclic Alternating Pattern

Abstract Introduction Upper Airway Resistance Syndrome (UARS) is suspected in individuals with excessive daytime sleepiness, fatigue, and sleep fragmentation associated with increased respiratory effort. UARS can negatively impact daytime function. Conventional polysomnography parameters do not demonstrate significant abnormalities in UARS patients but increase in RERAs and arousal index. Cyclic alternating pattern (CAP) is a periodic electroencephalogram activity of non-REM sleep that expresses a condition of sleep instability. The objective of the study was to compare CAP components between UARS patients and health individuals. Methods Fifteen subjects with UARS and 15 age- and sex- matched controls had their sleep study blinded analyzed. UARS criteria were the presence of sleepiness (Epworth Sleepiness Scale – ESS - ≥ 10) and/or fatigue (Modified Fatigue Impact Scale ≥ 38) associated with an apnea/hypopnea index (AHI) ≤ 5 and a respiratory disturbance index (RDI) &amp;gt; 5 events/hour of sleep, and/or flow limitation in more than 30% of total sleep time. Control group criteria were AHI &amp;lt; 5 events/hour, RDI ≤ 5 events/hour and &amp;lt; 30% of TST with flow limitation and ESS &amp;lt; 10, without sleep, clinical, neurological, or psychiatric disorder. CAP electroencephalogram of both groups was analyzed. Results We found higher CAP rate (p = 0.05) and CAP index in N1 stage (p &amp;lt; 0.001) and in N3 stage (p &amp;lt; 0.001) in UARS patients compared to control group. Considering only CAP phase A1 analysis, UARS patients presented higher CAP rate (p = 0.05) and CAP index in N1 stage (p &amp;lt; 0.001) as well as CAP index in N3 stage (p &amp;lt; 0.001) compared to control group. Considering only CAP phase A2 analysis, UARS patients also presented higher number of CAP in N1 stage (p = 0.05). There was no significant difference for CAP phase A3 between groups. Conclusion Although UARS is associated with high arousal index, we found increase in CAP phase A1 and A2, which do not include necessarily AASM arousals, suggesting not only sleep fragmentation but also sleep instability. Support (if any) Associação Fundo de Incentivo à Pesquisa (AFIP) and Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP).