Classification and automatic scoring of arousal intensity during sleep stages using machine learning.

Arousal during sleep can result in sleep fragmentation and various physiological effects, impairing cognitive function and raising blood pressure and heart rate. However, the current definition of arousal has limitations in assessing both amplitude and duration, making it challenging to measure sleep fragmentation accurately. Moreover, there is inconsistency among inter-raters in arousal scoring, which renders it susceptible to subjective variability. Therefore, this study aims to identify a highly accurate classifier for each sleep stage by employing optimized feature selection and machine learning models. According to electroencephalography (EEG) signals during the arousal phase, the intensity level was categorized into four levels. For control, the non-arousal cases were used as level 0 and referred as sham arousal, resulting in five arousal intensity levels. Wavelet transform was applied to analyze sleep arousal to extract features from EEG. Based on these features, we classified arousal intensity levels through machine learning algorithms. Due to the different characteristics of EEG in each sleep stage, the classification model was optimized for the four sleep stages. Excluding sham arousals, a total of 13,532 arousal events were used. The lowest intensity in the entire data, level 1, was computed to be 3107, level 2 was 3384, level 3 was 3472, and the highest intensity of level 4 was 3,569. The optimized classification model for each sleep stage achieved an average sensitivity of 82.68%, specificity of 95.68%, and AUROC of 96.30%. The sensitivity of the control, arousal intensity level 0, was 83.07%, a 1.25% increase over the unoptimized model and a 14.22% increase over previous research. This study used machine learning techniques to develop classifiers for each sleep stage, improving the accuracy of arousal intensity classification. The classifiers showed high sensitivity and specificity and revealed the unique characteristics of arousal intensity during different sleep stages. These findings represent a novel approach to arousal research and have implications for developing more accurate predictive models in sleep research.

The simplification of the insomnia severity index and epworth sleepiness scale using machine learning models.

Insomnia and excessive daytime sleepiness (EDS) are the most common complaints in sleep clinics, and the cost of healthcare services associated with them have also increased significantly. Though the brief questionnaires such as the Insomnia Severity Index (ISI) and Epworth Sleepiness Scale (ESS) can be useful to assess insomnia and EDS, there are some limitations to apply for large numbers of patients. As the researches using the Internet of Things technology become more common, the need for the simplification of sleep questionnaires has been also growing. We aimed to simplify ISI and ESS using machine learning algorithms and deep neural networks with attention models. The medical records of 1,241 patients who examined polysomnography for insomnia or EDS were analyzed. All patients are classified into five groups according to the severity of insomnia and EDS. To develop the model, six machine learning algorithms were firstly applied. After going through normalization, the process with the CNN+ Attention model was applied. We classified a group with an accuracy of 93% even with only the results of 6 items (ISI1a, ISI1b, ISI3, ISI5, ESS4, ESS7). We simplified the sleep questionnaires with maintaining high accuracy by using machine learning models.

Obstructive Sleep Apnea and Its Influence on Intracranial Aneurysm.

Obstructive sleep apnea syndrome (OSAS) is associated with cerebrovascular disease, which can lead to life-threatening outcomes. The purpose of the study was to investigate the relationship between OSAS and comorbid intracranial aneurysms. We retrospectively reviewed 564 patients who underwent a polysomnography and brain magnetic resonance angiography as part of their health checkup. We calculated the prevalence of an intracranial aneurysm and OSAS in patients and measured the size of the intracranial aneurysm if present. The mean patient age was 55.6 ± 8.5 years, and 82.3% of them were men. The prevalence of an intracranial aneurysm in patients with OSAS was 12.1%, which is significantly higher than patients with non-OSAS (5.9%, p = 0.031). Patients with OSAS had a much higher prevalence of intracranial aneurysms, after adjusting all possible confounding factors such as age, sex, smoking status, alcohol drinking, and body mass index (odds ratio: 2.32; 95% confidence interval: 1.07-5.04). Additionally, the OSAS group had noticeably larger aneurysms compared with those of the non-OSAS group (3.2 ± 2.0 mm vs. 2.0 ± 0.4 mm, p = 0.013). We found a significant association between OSAS and intracranial aneurysms. OSAS could be another risk factor for the development of intracranial aneurysms.

Effects of Obstructive Sleep Apnea on Epilepsy, and Continuous Positive Airway Pressure as a Treatment Option.

Purpose: To compare the characteristics of obstructive sleep apnea (OSA) between patients with epilepsy and patients without epilepsy and to investigate CPAP (Continuous Positive Airway Pressure) effect on seizures. Methods: Medical and polysomnography (PSG) data from 235 adult OSA patients with epilepsy (OE; 183 males; mean age, 49.8 years) and 268 age- and sex-matched OSA patients without epilepsy (OSE; 216 males; mean age, 51.3 years), obtained between March 2014 and May 2020 and housed in a database in a university-affiliated hospital, were retrospectively reviewed. All subjects completed surveys addressing comorbidities and medications, and sleep-related questionnaires including the Insomnia Severity Index, Epworth Sleepiness Scale, Pittsburgh Sleep Quality Index, and Beck Depression Inventory-II. Results: Compared with the OSE group, the OE group reported fewer sleep-related complaints and less severe OSA-related PSG parameters, with a lower apnea-hypopnea index (24.9 vs. 33.4 events/h; p < 0.003), arousal index (23.3 vs. 30.8 events/h; p < 0.001), and oxygen desaturation index (19.6 vs. 28.8; p < 0.002). The OE group had fewer smokers and lower alcohol consumption but a higher body mass index (27.0 vs. 25.9 kg/m2; p < 0.001). No correlations were observed between OSA-related PSG parameters and epilepsy-related factors, such as age at seizure onset, seizure type, frequency of seizures, presence of nocturnal seizures, and number of antiseizure medications, in the OE group. Patients with OE who demonstrated good compliance with CPAP therapy exhibited a decrease in seizure frequency. Conclusions: The OE group exhibited less severe disease characteristics than their age- and sex-matched OSE counterparts. Nevertheless, because the coexistence of OSA and epilepsy is high, CPAP therapy can reduce the frequency of seizures. Therefore, it is important to evaluate the presence of OSA in patients with epilepsy and to treat the conditions concurrently.

Sleep Disturbances and Phenoconversion in Patients with REM Sleep Behavior Disorder.

OBJECTIVE: We aimed to investigate relationships between sleep disturbances and phenoconversion to neurodegenerative diseases in patients with REM sleep behavior disorder (RBD). METHOD: Using a comprehensive sleep database in a university-affiliated hospital between December 2014 and March 2021, we reviewed the data of 226 patients with RBD (182 patients with idiopathic RBD (iRBD) and 44 patients with symptomatic RBD (sRBD) with a neurodegenerative disease). RESULTS: Among 226 patients with RBD (male, 61.5%), the mean age at RBD onset and mean disease duration were 59.4 ± 10.5 and 5.9 ± 5.6 years, respectively. Further, 111 (49.1%) patients had periodic limb movements during sleep (PLMS, PLM index ≥ 15/h), while 110 patients (48.7%) had comorbid obstructive sleep apnea (OSA, respiratory disturbance index ≥ 15/h). There was a positive correlation between age at RBD onset and the apnea-hypopnea index and Pittsburgh Sleep Quality Index. Compared to patients with iRBD, patients with sRBD showed a lower N3 sleep (3.3 ± 5.0 vs. 1.6 ± 3.1%, p = 0.004) and higher periodic limb movement index (36.3 ± 31.8 vs. 56.9 ± 47.5/h, p = 0.021) at the baseline. Among the 186 patients with iRBD, 18 (8.0%) developed neurodegenerative diseases (converters, mean follow-up duration: 2.5 ± 1.6 years) and 164 did not (non-converters, mean follow-up 2.4 ± 2.2 years). There was no significant between-group difference in the demographics and baseline clinical features. Continuous positive airway pressure (CPAP) therapy was prescribed in 101 patients with OSA; among them, 71 (70%) patients agreed to use it. CPAP improved dream enactment behaviors. CONCLUSION: In our study, 8.0% of patients with iRBD showed phenoconversion within a mean follow-up duration of 2.5 years. Polysomnographic parameters could not predict phenoconversion to neurodegenerative disease. However, approximately half of the patients with RBD presented with significant sleep disorders, including OSA or PLMS. CPAP therapy may alleviate RBD symptoms in patients with RBD-OSA.

Automated Detection of Obstructive Sleep Apnea Events from a Single-Lead Electrocardiogram Using a Convolutional Neural Network.

In this study, we propose a method for the automated detection of obstructive sleep apnea (OSA) from a single-lead electrocardiogram (ECG) using a convolutional neural network (CNN). A CNN model was designed with six optimized convolution layers including activation, pooling, and dropout layers. One-dimensional (1D) convolution, rectified linear units (ReLU), and max pooling were applied to the convolution, activation, and pooling layers, respectively. For training and evaluation of the CNN model, a single-lead ECG dataset was collected from 82 subjects with OSA and was divided into training (including data from 63 patients with 34,281 events) and testing (including data from 19 patients with 8571 events) datasets. Using this CNN model, a precision of 0.99%, a recall of 0.99%, and an F1-score of 0.99% were attained with the training dataset; these values were all 0.96% when the CNN was applied to the testing dataset. These results show that the proposed CNN model can be used to detect OSA accurately on the basis of a single-lead ECG. Ultimately, this CNN model may be used as a screening tool for those suspected to suffer from OSA.

Cardiopulmonary Coupling Analysis Using Home Sleep Monitoring System Based on Air Mattress.

This study investigates the feasibility of cardiopulmonary coupling (CPC) using home sleep monitoring system. We have designed a system to measure respiratory signals and normal-to-normal (NN) interval series in a non-contact based on air mattress. Then, CPC analysis was conducted using extracted respiratory signals and NN interval series, and six CPC parameters were extracted (VLFC, LFC, HFC, e-LFC, e-LFCNB and e-LFCBB). To evaluate the proposed method, two statistical analyses were conducted between the CPC parameters extracted by the electrocardiogram-based conventional method and the air mattress-based proposed method for five patients with obstructive sleep apnea and hypopnea (OSAH). Wilcoxon's signed rank test on the CPC parameters of the two methods indicated no significant differences (p > 0.05) and Spearman's rank correlation analysis showed high positive correlations (r ≥ 0.7, p < 0.05) between the two methods. Therefore, the proposed method has the potential for performing CPC analysis using air mattress-based system.

Estimating sleep parameters using nasal pressure signals applicable to continuous positive airway pressure devices.

OBJECTIVE: This paper proposes a method for classifying sleep-wakefulness and estimating sleep parameters using nasal pressure signals applicable to a continuous positive airway pressure (CPAP) device. APPROACH: In order to classify the sleep-wakefulness states of patients with sleep-disordered breathing (SDB), apnea-hypopnea and snoring events are first detected. Epochs detected as SDB are classified as sleep, and time-domain- and frequency-domain-based features are extracted from the epochs that are detected as normal breathing. Subsequently, sleep-wakefulness is classified using a support vector machine (SVM) classifier in the normal breathing epoch. Finally, four sleep parameters-sleep onset, wake after sleep onset, total sleep time and sleep efficiency-are estimated based on the classified sleep-wakefulness. In order to develop and test the algorithm, 110 patients diagnosed with SDB participated in this study. Ninety of the subjects underwent full-night polysomnography (PSG) and twenty underwent split-night PSG. The subjects were divided into 50 patients of a training set (full/split: 42/8), 30 of a validation set (full/split: 24/6) and 30 of a test set (full/split: 24/6). MAIN RESULTS: In the experiments conducted, sleep-wakefulness classification accuracy was found to be 83.2% in the test set, compared with the PSG scoring results of clinical experts. Furthermore, all four sleep parameters showed higher correlations than the results obtained via PSG (r ⩾ 0.84, p < 0.05). In order to determine whether the proposed method is applicable to CPAP, sleep-wakefulness classification performances were evaluated for each CPAP in the split-night PSG data. The results indicate that the accuracy and sensitivity of sleep-wakefulness classification by CPAP variation shows no statistically significant difference (p < 0.05). SIGNIFICANCE: The contributions made in this study are applicable to the automatic classification of sleep-wakefulness states in CPAP devices and evaluation of the quality of sleep.

Outdoor artificial light at night, obesity, and sleep health: Cross-sectional analysis in the KoGES study.

Obesity is a common disorder with many complications. Although chronodisruption plays a role in obesity, few epidemiological studies have investigated the association between artificial light at night (ALAN) and obesity. Since sleep health is related to both obesity and ALAN, we investigated the association between outdoor ALAN and obesity after adjusting for sleep health. We also investigated the association between outdoor ALAN and sleep health. This cross-sectional survey included 8526 adults, 39-70 years of age, who participated in the Korean Genome and Epidemiology Study. Outdoor ALAN data were obtained from satellite images provided by the US Defense Meteorological Satellite Program. We obtained individual data regarding outdoor ALAN; body mass index; depression; and sleep health including sleep duration, mid-sleep time, and insomnia; and other demographic data including age, sex, educational level, type of residential building, monthly household income, alcohol consumption, smoking status and consumption of caffeine or alcohol before sleep. A logistic regression model was used to investigate the association between outdoor ALAN and obesity. The prevalence of obesity differed significantly according to sex (women 47% versus men 39%, p < 0.001) and outdoor ALAN (high 55% versus low 40%, p < 0.001). Univariate logistic regression analysis revealed a significant association between high outdoor ALAN and obesity (odds ratio [OR] 1.24, 95% confidence interval [CI] 1.14-1.35, p < 0.001). Furthermore, multivariate logistic regression analyses showed that high outdoor ALAN was significantly associated with obesity after adjusting for age and sex (OR 1.25, 95% CI 1.14-1.37, p < 0.001) and even after controlling for various other confounding factors including age, sex, educational level, type of residential building, monthly household income, alcohol consumption, smoking, consumption of caffeine or alcohol before sleep, delayed sleep pattern, short sleep duration and habitual snoring (OR 1.20, 95% CI 1.06-1.36, p = 0.003). The findings of our study provide epidemiological evidence that outdoor ALAN is significantly related to obesity.

Cortical Thinning and Altered Cortico-Cortical Structural Covariance of the Default Mode Network in Patients with Persistent Insomnia Symptoms.

STUDY OBJECTIVES: Recent studies have suggested that structural abnormalities in insomnia may be linked with alterations in the default-mode network (DMN). This study compared cortical thickness and structural connectivity linked to the DMN in patients with persistent insomnia (PI) and good sleepers (GS). METHODS: The current study used a clinical subsample from the longitudinal community-based Korean Genome and Epidemiology Study (KoGES). Cortical thickness and structural connectivity linked to the DMN in patients with persistent insomnia symptoms (PIS; n = 57) were compared to good sleepers (GS; n = 40). All participants underwent MRI acquisition. Based on literature review, we selected cortical regions corresponding to the DMN. A seed-based structural covariance analysis measured cortical thickness correlation between each seed region of the DMN and other cortical areas. Association of cortical thickness and covariance with sleep quality and neuropsychological assessments were further assessed. RESULTS: Compared to GS, cortical thinning was found in PIS in the anterior cingulate cortex, precentral cortex, and lateral prefrontal cortex. Decreased structural connectivity between anterior and posterior regions of the DMN was observed in the PIS group. Decreased structural covariance within the DMN was associated with higher PSQI scores. Cortical thinning in the lateral frontal lobe was related to poor performance in executive function in PIS. CONCLUSION: Disrupted structural covariance network in PIS might reflect malfunctioning of antero-posterior disconnection of the DMN during the wake to sleep transition that is commonly found during normal sleep. The observed structural network alteration may further implicate commonly observed sustained sleep difficulties and cognitive impairment in insomnia.

Effects of long-term treatment on brain volume in patients with obstructive sleep apnea syndrome.

We assessed structural brain damage in obstructive sleep apnea syndrome (OSA) patients (21 males) and the effects of long-term continuous positive airway pressure (CPAP) treatment (18.2 ± 12.4 months; 8-44 months) on brain structures and investigated the relationship between severity of OSA and effects of treatment. Using deformation-based morphometry to measure local volume changes, we identified widespread neocortical and cerebellar atrophy in untreated patients compared to controls (59 males; Cohen's D = 0.6; FDR < 0.05). Analysis of longitudinally scanned magnetic resonance imaging (MRI) scans both before and after treatment showed increased brain volume following treatment (FDR < 0.05). Volume increase was correlated with longer treatment in the cortical areas that largely overlapped with the initial atrophy. The areas overlying the hippocampal dentate gyrus and the cerebellar dentate nucleus displayed a volume increase after treatment. Patients with very severe OSA (AHI > 64) presented with prefrontal atrophy and displayed an additional volume increase in this area following treatment. Higher impairment of working memory in patients prior to treatment correlated with prefrontal volume increase after treatment. The large overlap between the initial brain damage and the extent of recovery after treatment suggests partial recovery of nonpermanent structural damage. Volume increases in the dentate gyrus and the dentate nucleus possibly likely indicate compensatory neurogenesis in response to diminishing oxidative stress. Such changes in other brain structures may explain gliosis, dendritic volume increase, or inflammation. This study provides neuroimaging evidence that revealed the positive effects of long-term CPAP treatment in patients with OSA.