Apnea Events Research Articles

Skeletal Muscle Oxygen Kinetics During Exercise In Adults With Obstructive Sleep Apnea.

Obstructive sleep apnea (OSA) is associated with persistent and progressive nighttime sympathetic nervous system (SNS) arousal strain in response to both hypopnea and apnea events. This repetitive strain of nighttime activation of the SNS may promote daytime hyperactivity, possibly limiting microvascular reactivity. PURPOSE: The purpose of this study was to examine the on-kinetic profile of muscle deoxygenation during sub-maximal walking in adults with OSA. METHODS: Twelve adults with OSA (age=48±10years, BMI=29±5kg/m2, Apnea–hypopnea index (AHI)=50±24) and 12 healthy non-OSA (NO) adults (42±8years, BMI=24±3kg/m2) completed two 6-minute bouts of submaximal exercise on a motorized treadmill, corresponding to 85% of anaerobic threshold. Using near-infrared spectroscopy (NIRS), concentration changes in deoxygenated hemoglobin-myoglobin (∆[HHb]) was measured continuously from the left lateral gastrocnemius muscle. The two bouts were averaged to form a single ∆[HHb] response profile per subject. Indices of ∆[HHb] on-kinetics include the time constant (τ), ∆[HHb] amplitude (∆[HHb]amp), and mean response time (MRT=time delay+τ). In addition, the transition constant (Kt=∆[HHb]amp/MRT) reflects the overall normalized rate of ∆[HHb] on-kinetics. Data were compared using age, BMI, gender, race, total physical activity and sleep duration as covariates for ANCOVA. RESULTS: Both τ and MRT did not differ between the groups (OSA:19.8±8.2s, NO:19.4±4.4s, p=0.835; and OSA:28.9±7.2s, NO:24.4±5.9s, p=0.515 respectively). However, ∆[HHb]amp and Kt were lower in OSA compared to NO (OSA:2.8 ±2a.u., NO:8.7±5.3a.u., p=0.011; and OSA:0.101±0.07a.u/s, NO:0.354±0.17a.u./s, p=0.002, respectively). Further, the multivariate regression analysis showed that AHI was a strong predictor of and was negatively associated with ∆[HHb]amp as well as Kt (p=0.015 and p=0.019 respectively). CONCLUSION: The lower normalized response rate of muscle deoxygenation may be reflective of impaired capability of the muscle to extract/utilize oxygen, which may contribute towards a decreased ability to sustain physical activity. Disease severity was also directly related to the normalized ∆[HHb] response rate, which has implications for physical activity participation in persons with OSA.

Identification of Apnea Events Using a Chest-Worn Monitor Compared to Laboratory-Based Polysomnography in Patients Suspected of Obstructive Sleep Apnea

Obstructive sleep apnea (OSA) is an under-diagnosed risk factor for several adverse health outcomes. The gold standard diagnostic test for OSA is laboratory-based polysomnography (PSG). Portable sl...

Measurement of respiratory effort in sleep by 3D camera and respiratory inductance plethysmography

Polysomnography systems used for the diagnosis of sleep respiratory disorders are comprised of multiple sensors, including abdomen and thorax respiratory inductance plethysmography (RIP) belts to record respiratory effort. However, RIP belts are known to be susceptible to signal loss. To resolve this, we utilized a contactless three-dimensional (3D) time-of-flight (TOF) camera to monitor respiratory effort. We aimed to show that respiratory effort monitoring can be achieved by 3D TOF camera recording instead of RIP belts. The use of RIP belt signals is twofold. Firstly, the signals are used to classify the apnea events into obstructive, central, and mixed. Additionally, the American Academy of Sleep Medicine (AASM) Scoring Manual recommends the scoring of apneas and hypopneas using RIPSum (the sum of the abdomen and thorax RIP signals) when the airflow sensors are unavailable. We therefore used the 3D effort signal to classify apneas and compared it to the RIP signal classification. Reduced effort events from RIP and 3D effort signals were compared to the apnea and hypopnea events. Furthermore, the changes in effort during the events were compared between the two effort signals. Classification by 3D effort signal performed well, with 80% accuracy. It worked best for central apneas, with an accuracy of 99%. There was a high correlation of r = 0.88 (r ≠ 0, p = 0.0001) between the 3D effort signal events and RIPSum events. There was also a significant correlation of 0.62 (r ≠ 0, p = 0.0001) between 3D effort signal and RIPSum in the decrease of effort during apnea and hypopnea events. We conclude that respiratory effort derived from a 3D TOF camera can be used as an alternative to RIP belts for scoring of apneas and hypopneas and classification of apneas.

Spectral properties of the respiratory signal during sleep apnea events: Obtrusive and unobtrusive measurements

People with obstructive sleep apnea hypopnea syndrome (OSAHS) are affected by disruption in normal breathing patterns during sleep. In the literature, it is common to find acquisition of Thoracic (THO) and abdominal (ABD) movements with piezo-electric bands included in a full polysomnography. These movements convey valuable information related to sleep apnea events, and for this reason, contactless methods, such as the Pressure Bed Sensor (PBS), have been developed to extract this information. The main goal of this study is to analyze apnea and hypopnea fluctuations based on the spectral analysis of nasal airflow measure (as a reference signal), thoraco–abdominal effort and PBS respiration signal. To this end, features from the respiratory spectrum such as entropy, Gaussian modeling and instantaneous frequency were computed. These spectral properties were evaluated in three windows for each sensor: control point (CP) which is a window randomly extracted for the sleep time without apnea event, before event (BE) a window before an apnea episode and during event (DE) a window during an apnea episode. Apnea and hypopnea events were analyzed separately. According to a database of seventeen subjects, DE windows showed significant differences with respect to the CP window in most of the computed indices for both apnea and hypopnea events for all sensors. Significant differences were also found when DE and BE windows were compared in the case of apnea for all the sensors. In conclusion, the analyzed spectral characteristics could be a good tool to detect apnea and hypopnea. Finally, PBS signal which is a unobtrusive sensor, maintains the spectral properties of the standard respiratory effort measurements, and the use of this sensor could be useful for the monitoring outside of a clinical environment, simplifying the acquisition process.

Time domain characterization for sleep apnea in oronasal airflow signal: a dynamic threshold classification approach

Objective: Apneas are the most common type of sleep-related breathing disorders; they cause patients to move from restorative sleep into inefficient sleep. The American Academy of Sleep Medicine (AASM) considers sleep apnea as a hidden health crisis that affects 29.4 million adults, costing the USA billions of dollars. Traditional detection methods of sleep apnea are achieved by human observation of the respiration signals. This introduces limitations in terms of access and efficiency of diagnostic sleep studies. However, alternative device technologies have limited diagnostic accuracy for detecting apnea events although many of the previous investigational algorithms are based on multiple physiological channel inputs. Guided by the AASM recommendations for sleep apnea diagnostics, this paper investigates time domain metrics to characterize changes in oronasal airflow respiration signals during the occurrence of apneic events. Approach: A new algorithm is developed to derive a respiratory baseline from the oronasal airflow signal in order to detect sleep apnea events using a dynamically adjusted threshold classification approach. To demonstrate our results, we use polysomnography data of patients with different apnea severity levels as reflected by their overnight apnea hypopnea index (AHI), including patients with mild apnea (5 AHI ), moderate apnea ( AHI ), and severe apnea (AHI ). Main results: Our results indicate the ability to characterize sleep apnea events in oronasal airflow signals using the proposed dynamic threshold classification approach. Overall, the new algorithm achieved a sensitivity of 80.0%, specificity of 88.7%, and an area under receiver operating characteristics curve of 0.844. Significance: The present results contribute a new approach for progressive detection of sleep apnea using an adaptive threshold that is dynamically adjusted with respect to the patient’s respiration baseline, making it potentially able to effectively generalize over patients with different apnea severity levels and longer monitoring periods.

Polysomnography in Preterm Infants with Bronchopulmonary Dysplasia for Monitoring Sleep-Disordered Breathing and Pulmonary Reserve

Bronchopulmonary dysplasia (BPD) has progressed over time into a syndrome with multifactorial aetiology and complex pathophysiology characterised by a developmental arrest of the alveolar and pulmonary vasculature compartments. BPD remains common in extremely preterm and very low birth weight infants. Maintaining appropriate oxygen blood levels in BPD infants may promote growth, reduce the risk of sudden infant death syndrome, and lower pulmonary artery pressures. There is no agreed approach on how to best titrate and wean home oxygen treatment in BPD infants. In BPD infants on home oxygen therapy, sleep-disordered breathing is common and appears to be central in origin. However, obstructive apnoea events are also more common in BPD infants. The increased frequency of central apnoea events during sleep in BPD infants has been shown to decline on low-flow oxygen treatment to levels observed in healthy infants. It is hypothesised that brief respiratory pauses in sleep could result in significant oxygen desaturations in BPD infants who have a decreased pulmonary reserve. Those events are scored as central apnoea in polysomnography and are prevented by oxygen treatment. Central apnoea events may also represent disrupted control of breathing secondary to altered chemosensitivity. Polysomnography may be of clinical value to monitor sleep-disordered breathing, assess pulmonary reserves, and titrate and wean oxygen in BPD infants. Considering the increasing number of extremely preterm and very low birth weight infants with BPD, we recommend prioritising the performance of well-designed studies to gather high-level evidence into the potential role of polysomnography in the management of prematurity- and BPD-associated long-term sequelae.

Dynamic changes in electroencephalogram spectral power with varying apnea duration in older adults.

Sleep apnea elicits brain and physiological changes and its duration varies across the night. This study investigates the changes in the relative powers in electroencephalogram (EEG) frequency bands before and at apnea termination and as a function of apnea duration. The analysis was performed on 30 sleep records (375 apnea events) of older adults diagnosed with sleep apnea. Power spectral analysis centered on two 10-s EEG epochs, before apnea termination (BAT) and after apnea termination (AAT), for each apnea event. The relative power changes in EEG frequency bands were compared with changes in apnea duration, defined as Short (between 10 and 20s), Moderate (between 20 and 30s) and Long (between 30 and 40s). A significant reduction in EEG relative powers for lower frequency bands of alpha and sigma were observed for the Long compared to the Moderate and Short apnea duration groups at BAT, and reduction in relative theta, alpha and sigma powers for the Long compared to the Moderate and Short groups at AAT. The proportion of apnea events showed a significantly decreased trend with increased apnea duration for non-rapid eye movement sleep but not rapid eye movement sleep. The proportion of central apnea events decreased with increased apnea duration, but not obstructive episodes. The findings suggest EEG arousal occurred both before and at apnea termination and these transient arousals were associated with a reduction in relative EEG powers of the low-frequency bands: theta, alpha and sigma. The clinical implication is that these transient EEG arousals, without awakenings, are protective of sleep. Further studies with large datasets and different age groups are recommended.

Chemoreflex failure and sleep-disordered breathing in familial dysautonomia: Implications for sudden death during sleep.

Familial dysautonomia (Riley-Day syndrome, hereditary sensory and autonomic neuropathy type III) is a rare autosomal recessive disease characterized by impaired development of primary sensory and autonomic neurons resulting in a severe neurological phenotype, which includes arterial baroreflex and chemoreflex failure with high frequency of sleep-disordered breathing and sudden death during sleep. Although a rare disease, familial dysautonomia represents a unique template to study the interactions between sleep-disordered breathing and abnormal chemo- and baroreflex function. In patients with familial dysautonomia, ventilatory responses to hypercapnia are reduced, and to hypoxia are almost absent. In response to hypoxia, these patients develop paradoxical hypoventilation, hypotension, bradycardia, and potentially, death. Impaired ventilatory control due to chemoreflex failure acquires special relevance during sleep when conscious control of respiration withdraws. Overall, almost all adult (85%) and pediatric (95%) patients have some degree of sleep-disordered breathing. Obstructive apnea events are more frequent in adults, whereas central apnea events are more severe and frequent in children. The annual incidence rate of sudden death during sleep in patients with familial dysautonomia is 3.4 per 1000 person-year, compared to 0.5-1 per 1000 person-year of sudden unexpected death in epilepsy. This review summarizes recent developments in the understanding of sleep-disordered breathing in patients with familial dysautonomia, the risk factors for sudden death during sleep, and the specific interventions that could prevent it.

HEART-RATE BASED SLEEP APNEA DETECTION USING ARDUINO

The purpose of this study was to investigate the use of a cost-effective heart rate monitor sensor and Arduino Uno configuration to accurately detect simulated sleep apnea, through the use of the inter-beat interval (R-R interval). Three separate 30[Formula: see text]min heart rate recordings were taken, each with six simulated sleep apnea events ranging from 20 to 40[Formula: see text]s. The results were gathered and processed to identify the simulated sleep apnea events. In each of the recordings, the simulated sleep apnea events were visible and the key characteristics, surrounding the events, could be recognized. The heart rate monitor sensor and Arduino Uno configuration successfully detected the simulated sleep apnea events through the analysis and processing of the hearts R-R interval.

Obstructive sleep apnea phenotypes in men based on characteristics of respiratory events during polysomnography.

The upper airway (UA) anatomical collapsibility, UA muscle responsiveness, breathing control, and/or arousability are important contributing factors for obstructive sleep apnea (OSA). Differences in clinical manifestations of OSA are believed to reflect interactions among these factors. We aimed to classify OSA patients into subgroups based on polysomnographic (PSG) variables using cluster analysis and assess each subgroup's characteristics. Men with moderate or severe OSA and without any concomitant heart or psychosomatic disease were recruited. A hierarchical cluster analysis was performed using variables including fraction of apnea, respiratory event duration, minimum oxygen saturation, arousal rate before termination, and frequency of respiratory events in the supine position. The impact of sleep stages or body position on PSG variables was also evaluated in each cluster. A total of 210 men (mean age, 50.0years, mean body mass index, 27.4kg/m2) were studied. The three subgroups that emerged from the analysis were defined as follows: cluster 1 (high fraction of apnea and severe desaturation (20%)), cluster 2 (high fraction of apnea and long event duration (31%)), and cluster 3 (low fraction of apnea (49%)). There were differences in the body mass index and apnea type between the three clusters. Sleep stages and/or body position affected PSG variables in each cluster. Patients with OSA could be divided into three distinct subgroups based on PSG variables. This clustering may be used for assessing the pathophysiology of OSA to tailor individual treatment other than continuous positive airway pressure therapy.

Apnea hypopnea indices categorized by REM/NREM sleep and sleep positions in 100 children with adenotonsillectomy for obstructive sleep apnea disease

ObjectivesIn pediatric obstructive sleep apnea (OSA), the relationship between rapid eye movement sleep and upper airway collapse, and between sleep position and airway dimensions are well known. However, the interrelations between these factors and the obstructive apnea hypopnea index (O-AHI) have not been thoroughly investigated. MethodsA retrospective study including 100 children who underwent adenotonsillectomy between March 2010 and July 2017. Total O-AHI was divided into four subcategories by sleep stage and position. ResultsPreoperatively 14 of 47 mild cases of OSA (1 ≤ total O-AHI) and 17 of 18 moderate (5 ≤ total O-AHI) had time showing severe apnea (10 ≤ subcategorized O-AHI). Twenty-two of 24 severe cases (10 ≤ total O-AHI) exhibited very severe apnea (30 ≤ subcategorized O-AHI). All 11 very severe cases (30 ≤ total O-AHI) experienced more than 50 apnea events per hour in at least one of the O-AHI subcategories. After surgery, 23 of 70 cases classified as completely resolved (total O-AHI < 1) still had mild apnea in the O-AHI subcategories, and six of 13 cases who continued to experience apnea events had moderate-to-severe apnea. Seventeen cases worsened in the O-AHI subcategories, and total O-AHI deteriorated in two cases of the 17. The amount of REM sleep and use of the supine position increased significantly postoperatively in the quartile groups with the lowest baseline values (p < 0.0001). ConclusionsWhen an unexpected AHI value is encountered, the O-AHI subcategories may be informative regarding the indications for surgery and evaluating the efficacy thereof.

Automated Sleep Apnea Detection in Raw Respiratory Signals Using Long Short-Term Memory Neural Networks.

Sleep apnea is one of the most common sleep disorders and the consequences of undiagnosed sleep apnea can be very severe, ranging from increased blood pressure to heart failure. However, many people are often unaware of their condition. The gold standard for diagnosing sleep apnea is an overnight polysomnography in a dedicated sleep laboratory. Yet, these tests are expensive and beds are limited as trained staff needs to analyze the entire recording. An automated detection method would allow a faster diagnosis and more patients to be analyzed. Most algorithms for automated sleep apnea detection use a set of human-engineered features, potentially missing important sleep apnea markers. In this paper, we present an algorithm based on state-of-the-art deep learning models for automatically extracting features and detecting sleep apnea events in respiratory signals. The algorithm is evaluated on the Sleep-Heart-Health-Study-1 dataset and provides per-epoch sensitivity and specificity scores comparable to the state of the art. Furthermore, when these predictions are mapped to the apnea-hypopnea index, a considerable improvement in per-patient scoring is achieved over conventional methods. This paper presents a powerful aid for trained staff to quickly diagnose sleep apnea.

An IoMT cloud-based real time sleep apnea detection scheme by using the SpO2 estimation supported by heart rate variability

Abstract Obstructive sleep apnea refers to a highly rampant sleep-related breathing disorder. The gold standard examination for diagnosis is polysomnography. Even though it provides highly accurate results, this multi-parametric test is time consuming and expensive. It also does not align with the new trend in health care, where focus is shifted to wellness and prevention. One possible way to address this problem is home health care, through the use of minimal invasive devices, higher accessibility, and provision of low cost diagnosis. To manage this, an automated and portable sleep apnea detector was formulated and assessed. The device utilizes one SpO2 sensor for estimating the heart rate and the oxygen blood level as well. The basis of the proposed analysis method is the connection between heart rate variability and oxygen saturation with d apnea events. The measured signals were then transferred to a cloud-based system architecture to diagnose and warn the remote patients. This solution was used to process the data and display it on both the mobile phone and personal computer. Testing of the proposed algorithms was done using the St. Vincents University Hospital/University College Dublin sleep apnea database. Apart from this database, the researchers utilized data gathered from 10 apnea patient volunteers. The performance of the proposed scheme algorithm achieved an average accuracy, specificity, and sensitivity of 98.54, 98.95%, and 97.05%, respectively.

Reduction of sleep-disordered breathing following effective percutaneous mitral valve repair with the MitraClip system.

This study tested the hypothesis that a reduction of pulmonary congestion achieved by a reduction of mitral regurgitation (MR) severity in heart failure (HF) patients is associated with reduced event lengths of sleep-disordered breathing (SDB). We prospectively enrolled 20 consecutive HF patients who underwent MitraClip implantation. Patients underwent cardiorespiratory polygraphic recording prior to and after percutaneous mitral valve repair (PMVR). Beyond routinely established indicators of apneas and hypopneas per hour (respiratory event index), we manually analyzed apnea event lengths. MitraClip implantation led to marked reduction of MR severity and a reduction in left atrial pressure. These hemodynamic changes were accompanied by changes in SDB: the subtype of SDB switched from CSA to OSA in 4 patients. Likewise, quantitative indicators of SDB were altered in both forms of SDB with a reduction in circulatory delay (CSA 38 ± 14 vs. 33 ± 15s.; p = 0.002 and OSA 34 ± 9 vs. 28 ± 6s.; p = 0.02) and a corresponding reduction in ventilation lengths in CSA patients (42 ± 15 vs. 37 ± 13s.; p = 0.05). A reduction of pulmonary congestion as achieved by a decrease of left atrial pressure through successful MitraClip implantation is associated with a reduction in respiratory event lengths, further pointing towards a relation between SDB and HF.

Erythromycin and Reflux Events in Premature Neonates: A Randomized Clinical Trial.

Gastroesophageal reflux disease (GERD) in premature neonates may manifest as apnea, bradycardia, growth failure, aspiration, or feeding intolerance. Erythromycin ethylsuccinate (EES), is often used as a pro-kinetic in the management of GERD, despite lack of evidence or safety from randomized controlled trials. We sought to study the efficacy of enteral EES at a dose of 50 mg · kg · day in decreasing the frequency of gastroesophageal reflux events as determined by pH-multichannel intraluminal impedance (pH-MII) monitoring. In a randomized, double-blind, placebo-controlled trial, eligible premature neonates with clinical signs of GERD underwent 24-hour pH-MII monitoring. If >5 reflux events were identified on pH-MII, then subjects were randomized to receive either EES or placebo. Repeat 24-hour pH-MII was performed on day 7 of study treatment and compared to initial pH-MII. Forty-three premature neonates were enrolled. Of those, 31 neonates were randomized, 15 to EES and 16 to placebo with a median (IQR) pretreatment total reflux events per 24 hours of 23 (16-40) and 29 (12-40), respectively. Day 7 total events per 24 hours decreased by 4 events in the EES group to 19 (15-33) and by 10 events in the placebo group to 19 (11-26) (P = 0.09). There were no differences in pretreatment and day 7 acidic and nonacidic reflux, proximal reflux, total or percent reflux time, median or longest bolus clearance time, or nurse-reported apnea events between groups. Enteral EES did not decrease reflux events on 24-hour pH-MII at the dose studied. Therefore, it may be ineffective in the treatment of GERD in premature neonates.

Upper Airway Narrowing during Central Apnea in Obese Adolescents.

The use of real-time magnetic resonance imaging (MRI) for the evaluation during sleep-related respiratory events can lead to better understanding of airway dynamics. To investigate the dynamic anatomy of the upper airway during central apnea. The study included obese adolescents who snore and were otherwise healthy. Subjects underwent an overnight baseline polysomnogram. Subjects slept during a 24-minute real-time upper airway MRI scan wearing a full face mask attached to a pneumotach. Sleep versus wakefulness during the MRI was inferred from the heart rate and respiratory patterns. Central apneas were scored using tracings of facemask airflow and abdominal bellows. The cross-sectional area of the upper airway before, during, and after each central apnea event was recorded. Eight subjects were studied and 57 central apnea events were observed during real-time MRI scanning during natural sleep. The median age of subjects was 14.0 years (interquartile range [IQR], 13.5 to 15.5). The median average reduction in cross-sectional area during central apnea events was -38% (IQR, -27 to -51) for primary snorers and -45% (IQR, -40 to -54) for subjects with obstructive sleep apnea. The percentage decrease in cross-sectional area of upper airway during a central apnea event was positively correlated to the length of the central apnea (ρ = 0.389; r2 = 0.152; P = 0.003). We observed that there is upper airway narrowing during central apneas during natural sleep in obese adolescent subjects, using real-time MRI.

Utility of transthoracic impedance and novel algorithm for sleep apnea screening in pacemaker patient

BackgroundPacing patients was revealed with a high prevalence of sleep disorder, but mostly undiagnosed. The pacemaker with transthoracic impedance sensor and novel algorithm could identify sleep apnea (SA) event. This study aimed to evaluate accuracy of pacemaker in sleep apnea diagnosis.MethodsThis is a prospective study which enrolled patients implanted with pacemakers integrated with transthoracic impedance sensor and SA-identified algorithm (AP Scan). All patients underwent a polysomnography (PSG). The apnea and hypopnea index (AHI) of the PSG (PSG-AHI) and the respiratory disturbance index (RDI) of the pacemaker (PM-RDI) were recorded on the same night. The correlation between two methods was evaluated by the kappa coefficient, receiver operating characteristic (ROC) curves, and Bland and Altman statistics.ResultsSixty-four patients were enrolled, who had never been diagnosed with SAS or underwent PSG exam. After PSG examination, 76.4% patients were diagnosed as combining with SA (20% severe, 18.2% moderate, and 38.2% mild). RDI calculated by PM has a strong positive correlation with PSG-AHI (r = 0.76, p < 0.001, 95% CI 0.61–0.85). The optimal cutoff value of PM-RDI for advanced SAS (PSG-AHI ≥ 15) diagnosis was 26, with AUC of 0.89 (95% CI 0.77 to 0.96, p < 0.001). The best cutoff value for severe SA (PSG-AHI ≥ 30) identification was 41, with a sensitivity of 81.6%, a specificity of 88.6%.ConclusionsPacemaker patients present a high prevalence of undiagnosed SA. Detection of SA by pacemaker is feasible and accurate in SA screening and monitoring.

Expert-level sleep scoring with deep neural networks.

ObjectivesScoring laboratory polysomnography (PSG) data remains a manual task of visually annotating 3 primary categories: sleep stages, sleep disordered breathing, and limb movements. Attempts to automate this process have been hampered by the complexity of PSG signals and physiological heterogeneity between patients. Deep neural networks, which have recently achieved expert-level performance for other complex medical tasks, are ideally suited to PSG scoring, given sufficient training data.MethodsWe used a combination of deep recurrent and convolutional neural networks (RCNN) for supervised learning of clinical labels designating sleep stages, sleep apnea events, and limb movements. The data for testing and training were derived from 10 000 clinical PSGs and 5804 research PSGs.ResultsWhen trained on the clinical dataset, the RCNN reproduces PSG diagnostic scoring for sleep staging, sleep apnea, and limb movements with accuracies of 87.6%, 88.2% and 84.7% on held-out test data, a level of performance comparable to human experts. The RCNN model performs equally well when tested on the independent research PSG database. Only small reductions in accuracy were noted when training on limited channels to mimic at-home monitoring devices: frontal leads only for sleep staging, and thoracic belt signals only for the apnea-hypopnea index.ConclusionsBy creating accurate deep learning models for sleep scoring, our work opens the path toward broader and more timely access to sleep diagnostics. Accurate scoring automation can improve the utility and efficiency of in-lab and at-home approaches to sleep diagnostics, potentially extending the reach of sleep expertise beyond specialty clinics.

Improvement in Nocturnal Hypoxemia in Obese Patients with Obstructive Sleep Apnea after Bariatric Surgery: a Meta-Analysis

To conduct a meta-analysis examining the effects of bariatric surgery on nocturnal hypoxemia in obese patients with obstructive sleep apnea (OSA). PubMed, EMBASE, Cochrane Library, and Web of Science were searched (the last search date was June 10, 2018) to identify relevant clinical studies. The mean arterial oxygen saturation (MeanSaO2), nadir oxygen saturation (NadirSaO2), apnea hypopnea index (AHI), and body mass index (BMI) data during the perioperative period were extracted and analyzed using a random effects model. Then, we performed subgroup and sensitivity analyses and calculated the publication bias to assess the between-study heterogeneity. In total, 15 studies with 636 patients were included; 13 were prospective observational trials, 1 was a randomized controlled trial (RCT), and 1 was a retrospective trial. After surgery, the MeanSaO2 and NadirSaO2 increased by 1.36 [95% CI (0.72, 2.00)] and 1.08 [95% CI (0.68, 1.49)], respectively, and the AHI and BMI decreased by 1.11 [95% CI (0.82, 1.40)] and 1.97 [95% CI (1.67, 2.27)], respectively. However, the heterogeneity across all trials was high; we identified some of the sources of that heterogeneity through subsequent subgroup and sensitivity analyses. Bariatric surgery is effective at improving nocturnal hypoxemia in obese patients with OSA; it also reduces body weight and the number of apnea events. More randomized controlled and comparative trials are necessary in the future to confirm our findings and to explore the potential underlying mechanisms.

Differences in arousal probability and duration after apnea and hypopnea events in adult obstructive sleep apnea patients

Objective: In obstructive sleep apnea (OSA), breathing cessations are often followed by arousals, leading to sleep fragmentation and thus impaired sleep quality. Arousals and fragmented sleep are also related to detrimental cardiovascular events. The key index for OSA diagnosis (i.e. the apnea–hypopnea index) attributes equal diagnostic value to apneas and hypopneas, despite the fact that the associated arousals and desaturations may be very different. Thus, considering the severity of the consequences of apneas and hypopneas could enhance the estimation of OSA severity. In this study, we investigate whether the probability and duration of apnea- and hypopnea-related arousals differ and whether the differences in desaturation severity following apneas and hypopneas are dependent on sleep stage. Approach: Polysomnographic recordings of 348 consecutive OSA patients were included for analysis. The severity of arousals and desaturations associated with hypopneas within different sleep stages was compared to that of arousals and desaturations associated with apneas. In addition, the probability of arousals related to apneas and hypopneas was evaluated within OSA severity categories. Main results: Apneas caused arousals less frequently than hypopneas in N1, N2, and N3 sleep in all OSA severity categories. However, the arousals caused by apneas were longer (p < 0.001) and the desaturations related to apneas were more severe (p < 0.001) than those related to hypopneas in N1, N2, and rapid eye movement sleep even after adjustment for respiratory event durations. Significance: Desaturations and arousals related to apneas are more severe than those related to hypopneas. Therefore, apneas followed by arousal or desaturation should have a different diagnostic value than hypopneas when assessing OSA severity and related risk for cardiovascular consequences.