Detection Of Obstructive Sleep Apnea Research Articles

Obstructive Sleep Apnea Syndrome Identification Using CNN-LSTM Hybrid Model

Obstructive Sleep Apnea (OSA) is a common sleep problem. It causes breathing issues during sleep. This disturbs oxygen intake and sleep patterns. Detecting OSA early and accurately is important for treatment. In our study, we look into using advanced deep learning for OSA detection. We focus on Long Short-Term Memory (LSTM) networks, Convolutional Neural Networks (CNN), and a new combined method. We also check the impact of different functions and methods on OSA detection accuracy. Our tests reveal some interesting facts. The Nadam optimizer gives the best OSA detection at 95.22% accuracy. The combined CNN and LSTM method achieves 95.38% accuracy. Using LSTM with the sigmoid function, we get a 93.80% accuracy. These results stress the importance of choosing suitable models and methods for OSA detection. Our study helps improve diagnostic tools for OSA. These findings can aid in early and improved treatment of OSA, giving people a better quality of life and health.

Automatic detection of obstructive sleep apnea based on speech or snoring sounds: a narrative review.

Obstructive sleep apnea (OSA) is a common chronic disorder characterized by repeated breathing pauses during sleep caused by upper airway narrowing or collapse. The gold standard for OSA diagnosis is the polysomnography test, which is time consuming, expensive, and invasive. In recent years, more cost-effective approaches for OSA detection based in predictive value of speech and snoring has emerged. In this paper, we offer a comprehensive summary of current research progress on the applications of speech or snoring sounds for the automatic detection of OSA and discuss the key challenges that need to be overcome for future research into this novel approach. PubMed, IEEE Xplore, and Web of Science databases were searched with related keywords. Literature published between 1989 and 2022 examining the potential of using speech or snoring sounds for automated OSA detection was reviewed. Speech and snoring sounds contain a large amount of information about OSA, and they have been extensively studied in the automatic screening of OSA. By importing features extracted from speech and snoring sounds into artificial intelligence models, clinicians can automatically screen for OSA. Features such as formant, linear prediction cepstral coefficients, mel-frequency cepstral coefficients, and artificial intelligence algorithms including support vector machines, Gaussian mixture model, and hidden Markov models have been extensively studied for the detection of OSA. Due to the significant advantages of noninvasive, low-cost, and contactless data collection, an automatic approach based on speech or snoring sounds seems to be a promising tool for the detection of OSA.

Severe obstructive sleep apnea in children with syndromic craniosynostosis: analysis of pulse transit time.

We examined the association between pulse transit time (PTT) and obstructive sleep apnea (OSA) in children with syndromic craniosynostosis (SCS), where OSA is a common problem and may cause cardiorespiratory disturbance. Retrospective study of children (age <18 years) with SCS and moderate-to-severe OSA (i.e., obstructive apnea-hypopnea index [oAHI] ≥ 5), or no OSA (oAHI < 1) who underwent overnight polysomnography (PSG). Children without SCS and normal PSG were included as controls. Reference intervals (RIs) for PTT were computed by non-parametric bootstrap analysis. Based on RIs of controls, the sensitivity and specificity of PTT to detect OSA were determined. In a linear mixed-model the explanatory variables assessed were sex, age, sleep stage, and time after obstructive events. In all 68 included children (19 SCS with OSA, 30 SCS without OSA, 19 controls), obstructive events occurred throughout all sleep stages, most prominently during rapid eye movement sleep (REM) and non-REM sleep stages N1 and N2, with evident PTT changes. Greatest reductions were observed 4 - 8 s after an event (p < 0.05). In SCS with OSA, PTT RIs were lower during all sleep stages compared to SCS without OSA. The highest sensitivity was observed during N1 (55.5%), and the highest specificity during REM (76.5%). Lowest PTT values were identified during N1. Obstructive events occur throughout all sleep stages with transient reductions in PTT. However, PTT as a variable for OSA detection is limited by its sensitivity and specificity.

Obstructive sleep apnea detection using optimized Bi-LSTM with random forest based exhaustive feature selector

Obstructive sleep apnea detection using optimized Bi-LSTM with random forest based exhaustive feature selector

The relationship between birth weight and obstructive sleep apnea in children

Objective:To study the relationship between children's birth weight and obstructive sleep apnea（OSA）. Methods:The sleep data and birth information of children who underwent polysomnography in the Department of Otorhinolaryngology-Head and Neck Surgery of Henan Children's Hospital from October 2020 to July 2022 were retrospectively analyzed. The data of OSA detection rate, OSA severity, sleep structure and respiratory parameters in different birth weight groups were analyzed. Results:A total of 2 778 children met the inclusion criteria, including 1 833 males and 945 females. According to birth weight, the selected children were divided into three groups: 122 small for gestational age（SGA） group, 2 313 appropriate for gestational age（AGA）, and 343 large for gestational age（LGA） group. There was no significant difference in age between different groups（P=0.061）. In each group, boys are significantly more numerous than girls（P=0.001）. The difference in current body mass index（BMI） between groups was statistically significant: the current BMI was higher in the LGA group（17.51±4.01, P<0.001）. The severity of OSA was different in different birth weight groups（P=0.037）. There was a strong positive correlation between the severity of OSA and birth weight（r=0.992）. Children in the SGA group had shorter rapid eye movement（REM） sleep period（19.00[15.18, 23.33], P=0.012）, higher obstructive apnea-hypopnea index（OAHI） values（1.75[0.60, 5.13], P=0.019）, and had lower central apnea hypopnea index（CAHI） values（0.10[0.00, 0.50], P=0.020）. There were no significant differences in sleep structure and respiratory parameters between the LGA group and the AGA group. Multiple regression analysis of the factors affecting the OAHI index showed that the OAHI index of boys was higher than that of girls（95%CI 1.311-2.096, P<0.001）, and age was negatively correlated with the OAHI index（r=-0.105, 95%CI 0.856-0.946, P<0.001）, current BMI and OAHI index were positively correlated（r=0.037, 95%CI 1.010-1.065, P=0.007）. LGA was positively correlated with OAHI index（r=0.346, 95%CI 1.039-1.921, P=0.027）, and the correlation between LGA and OAHI（r=0.346） was higher than that between SGA and OAHI（r=0.340）. Conclusion:There was no significant difference in the incidence of OSA in children with different birth weight groups, but the OSA severity of LGA group was higher. Gender, age, BMI index and large for gestational age were the influencing factors for the occurrence of OSA in children, which should be paid more attention to in clinical practice.

Association between risk of obstructive sleep apnea severity and risk of severe COVID-19 symptoms: insights from salivary and serum cytokines.

Obstructive sleep apnea (OSA) can adversely affect the immune response through clinical factors such as hypoxia, inflammation, and sleep disturbance. Since SARS-CoV-2 heavily relies on local and systemic host immune responses, this study aims to examine the links between the severity of OSA risk, cytokine levels, and the severity of symptoms associated with SARS-CoV-2 infection. Saliva and blood samples from 50 COVID-19 patients and 30 non-infected hospital staff members were collected. Using Luminex multiplex analysis, 65 blood and salivary cytokines were examined from the collected samples. Ordinal logistic regression analysis was utilized to examine the association between the self-reported risk of OSA, assessed through the STOP-Bang questionnaire, and the likelihood of experiencing severe symptoms of COVID-19. Mann-Whitney test was then performed to compare the cytokine levels between individuals with moderate to severe risk of OSA to those with a mild risk of OSA. Ordinal logistic regression analysis revealed that individuals with a moderate to severe risk of OSA were 7.60 times more likely to experience more severe symptoms of COVID-19 compared to those with a mild risk of OSA (OR = 7.60, 95%CI: 3.03, 19.06, p < 0.001). Moreover, among COVID-19-positive patients with a moderate to severe risk of OSA, there was a statistically significant negative correlation with serum IL-6 (p < 0.05), Eotaxin (CCL11) (p = 0.04), and salivary MIP-3α/CCL20 (p = 0.04). In contrast, individuals without COVID-19 who had a moderate to severe risk of OSA exhibited a significant positive correlation with serum IL-6 (p = 0.04). Individuals with moderate to severe risk of OSA were more likely to experience severe COVID-19 symptoms than those with mild risk for OSA. Additional analysis from the present studies revealed distinct patterns of oral and systemic immune responses between individuals with mild and moderate to severe risk of OSA. Findings from the present study underscores the importance of early detection and management of OSA to improve clinical outcomes, particularly when faced with the subsequent superimposed infection such as COVID-19.

Impact of obstructive sleep apnea on patients'’ quality of life

Abstract Obstructive sleep apnea (OSA) is a very commonly spread disease, not only in Bulgaria, but all over the world. One of the reasons for this is that patients suffering from OSA find it difficult to share their problem with their treating doctor (sleep doctor or ENT specialist). Doctor-patient communication is extremely important when gaining the patient'’s trust, for the early detection and diagnosis of obstructive sleep apnea. Sleep disorders lead to a decrease in the quality of life, due to frequent anxiety, fatigue, depressive states, reduced cognitive and motor functions, lack of energy, headaches and sleepiness throughout the day. Improving the quality of life of patients is a primary goal of the health care system, which is conveyed in patient care aimed at reducing or eliminating the effects of diseases and their associated negative psychological and social effects. Prevention, screening and effective communication with the patient are extremely important for a successful diagnosis and treatment of sleep disorders.

A bimodal feature fusion convolutional neural network for detecting obstructive sleep apnea/hypopnea from nasal airflow and oximetry signals

The most prevalent sleep-disordered breathing condition is Obstructive Sleep Apnea (OSA), which has been linked to various health consequences, including cardiovascular disease (CVD) and even sudden death. Therefore, early detection of OSA can effectively help patients prevent the diseases induced by it. However, many existing methods have low accuracy in detecting hypopnea events or even ignore them altogether. According to the guidelines provided by the American Academy of Sleep Medicine (AASM), two modal signals, namely nasal pressure airflow and pulse oxygen saturation (SpO2), offer significant advantages in detecting OSA, particularly hypopnea events. Inspired by this notion, we propose a bimodal feature fusion CNN model that primarily comprises of a dual-branch CNN module and a feature fusion module for the classification of 10-second-long segments of nasal pressure airflow and SpO2. Additionally, an Efficient Channel Attention mechanism (ECA) is incorporated into the second module to adaptively weight feature map of each channel for improving classification accuracy. Furthermore, we design an OSA Severity Assessment Framework (OSAF) to aid physicians in effectively diagnosing OSA severity. The performance of both the bimodal feature fusion CNN model and OSAF is demonstrated to be excellent through per-segment and per-patient experimental results, based on the evaluation of our method using two real-world datasets consisting of polysomnography (PSG) recordings from 450 subjects.

A fine-grained convolutional recurrent model for obstructive sleep apnea detection

A fine-grained convolutional recurrent model for obstructive sleep apnea detection

Validation of the Apnealink Air for diagnosis of obstructive sleep apnoea (OSA) in pregnant women in early-mid gestation.

The detection of obstructive sleep apnoea (OSA) in pregnant women in early-mid gestation is logistically difficult. Accurate alternates to polysomnography (PSG) in early pregnancy are not well identified. We compared the agreement between Apnealink Air (AL) and existing screening questionnaires to PSG in pregnant women ≤ 24-week gestation. Pregnant women (≤ 24-week gestation) underwent AL at home plus attended PSG in any order, completed within 7days where practicable. AL was manually scored (AL(M)) and automatically scored (AL(A)). An apnoea-hypopnea index (AHI) ≥ 5 was considered diagnostic of OSA and an AHI ≥ 15 considered at least moderate OSA. Diagnostic analysis was undertaken (sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV)) by generating receiver operating characteristic (ROC) curves and an area under the curve (AUC) (95% CI). Bland-Altman plots were used to plot agreement. Screening questionnaires (Epworth sleepiness score (ESS), STOP-BANG, calculated pregnancy-specific screening tool) were compared to PSG. A total of 49 participants successfully completed both tests at around 14-weeks gestation (IQR 12.9, 17.1). The time interval between AL and PSG was a median of 2days (IQR 1, 5 (range 1-11)). A total of 14 (29%) participants had OSA. The median AHI of AL(A) (3.1(IQR 0.85,4.6)) and AL(M) (IQR2.4(0.65,4.8)) did not differ from PSG (1.7(IQR1.0,6.1)). AL(A) and AL(M) compared to PSG demonstrated diagnostic test accuracy (area under curve (ROC)) of 0.94(95% CI 0.87-1.0) and 0.92(95% CI 0.85-1.0) respectively. Apnealink Air outperformed screening questionnaires tested. The findings suggest thatApnealink mayprovide a substitute to attended PSG identification of OSA in pregnant women in early-mid gestation using both manual and auto-scoring methods.

ECGAN-Assisted ResT-Net Based on Fuzziness for OSA Detection.

Growing attention has been paid recently to electrocardiogram (ECG) based obstructive sleep apnea (OSA) detection, with some progresses been made on this topic. However, the lack of data, low data quality, and incomplete data labeling hinder the application of deep learning to OSA detection, which in turn affects the overall generalization capacity of the network. To address these issues, we propose the ResT-ECGAN framework. It uses a one-dimensional generative adversarial network (ECGAN) for sample generation, and integrates it into ResTNet for OSA detection. ECGAN filters the generated ECG signals by incorporating the concept of fuzziness, effectively increasing the amount of high-quality data. ResT-Net not only alleviates the problems caused by deepening the network but also utilizes multihead attention mechanisms to parallelize sequence processing and extract more valuable OSA detection features by leveraging contextual information. Through extensive experiments, we verify that ECGAN can effectively improve the OSA detection performance of ResT-Net. Using only ResT-Net for detection, the accuracy on the Apnea-ECG and private databases is 0.885 and 0.837, respectively. By adding ECGAN-generated data augmentation, the accuracy is increased to 0.893 and 0.848, respectively. Comparing with the state-of-the-art deep learning methods, our method outperforms them in terms of accuracy. This study provides a new approach and solution to improve OSA detection in situations with limited labeled samples.

Detection of Obstructive Sleep Apnoea Using Features Extracted From Segmented Time-Series ECG Signals With a One Dimensional Convolutional Neural Network

Detection of Obstructive Sleep Apnoea Using Features Extracted From Segmented Time-Series ECG Signals With a One Dimensional Convolutional Neural Network

A Machine Learning Model for Obstructive Sleep Apnea Detection Using Ensemble Learning and Single-Lead EEG Signal Data

A Machine Learning Model for Obstructive Sleep Apnea Detection Using Ensemble Learning and Single-Lead EEG Signal Data

An Ensemble Learning-Assisted Obstructive Sleep Apnea Detection Model Using EEG Physiological Signals and Improved Extra Tree Classifier

An Ensemble Learning-Assisted Obstructive Sleep Apnea Detection Model Using EEG Physiological Signals and Improved Extra Tree Classifier

Retracted: A Comprehensive Review: Computational Models for Obstructive Sleep Apnea Detection in Biomedical Applications.

[This retracts the article DOI: 10.1155/2022/7242667.].

Advancements in Home-Based Devices for Detecting Obstructive Sleep Apnea: A Comprehensive Study

Obstructive Sleep Apnea (OSA) is a respiratory disorder characterized by frequent breathing pauses during sleep. The apnea–hypopnea index is a measure used to assess the severity of sleep apnea and the hourly rate of respiratory events. Despite numerous commercial devices available for apnea diagnosis and early detection, accessibility remains challenging for the general population, leading to lengthy wait times in sleep clinics. Consequently, research on monitoring and predicting OSA has surged. This comprehensive paper reviews devices, emphasizing distinctions among representative apnea devices and technologies for home detection of OSA. The collected articles are analyzed to present a clear discussion. Each article is evaluated according to diagnostic elements, the implemented automation level, and the derived level of evidence and quality rating. The findings indicate that the critical variables for monitoring sleep behavior include oxygen saturation (oximetry), body position, respiratory effort, and respiratory flow. Also, the prevalent trend is the development of level IV devices, measuring one or two signals and supported by prediction software. Noteworthy methods showcasing optimal results involve neural networks, deep learning, and regression modeling, achieving an accuracy of approximately 99%.

Anthropometric screening approach for obstructive sleep apnea in Japanese men: development and validation of the ABC scale

BackgroundThe existing screening tools for the detection of moderate and severe obstructive sleep apnea (OSA) are tailored to Western individuals. The aim of this study was to formulate and validate a simple anthropometric tool for OSA screening in a Japanese male population by incorporating objective indicators obtained during routine medical checkups.MethodsIn total, 443 Japanese men, who were suspected of having OSA and visited our sleep disorder clinic from 2009 to 2012, were retrospectively divided into two groups (development group [DG], n = 206; validation group [VG], n = 237). In the DG, OSA was defined as an apnea–hypopnea index of ≥ 15 events/h. Data regarding clinical and anthropometric variables (age, body mass index [BMI], neck circumference [NC], waist circumference [WC]) were obtained. For parameters that were associated with OSA on simple logistic regression analysis, cutoff values were calculated with receiver operating characteristic curves to create a new screening tool. The utility of the tool was evaluated in the VG and compared to that of the modified Berlin questionnaire (mBQ) for OSA.ResultsAge (odds ratio [OR] = 1.033, p = 0.002), BMI (OR = 1.231, p < 0.001), NC (OR = 1.327, p < 0.001), and WC (OR = 1.083, p < 0.001) were significant risk factors for OSA in the DG. Cutoff values for the prediction of OSA were age: ≥ 46.5 years, BMI: ≥ 25.45 kg/m2, NC: ≥ 38.35 cm, and WC: ≥ 90.75 cm. Our screening tool, incorporating age, BMI, and WC, yielded 70.1% sensitivity (mBQ: 85.4%) and 66.0% specificity (mBQ: 35.0%) in the VG.ConclusionsOSA screening in Japanese men has been conducted only in limited settings such as sleep clinics. The study findings support the implementation of screening in medical checkups using objective indices to enable early detection and treatment of OSA.

Wavelet transform and deep learning-based obstructive sleep apnea detection from single-lead ECG signals.

Sleep apnea is a common sleep disorder. Traditional testing and diagnosis heavily rely on the expertise of physicians, as well as analysis and statistical interpretation of extensive sleep testing data, resulting in time-consuming and labor-intensive processes. To address the problems of complex feature extraction, data imbalance, and low model capacity, we proposed an automatic sleep apnea classification model (CA-EfficientNet) based on the wavelet transform, a lightweight neural network, and a coordinated attention mechanism. The signal is converted into a time-frequency image by wavelet transform and put into the proposed model for classification. The effects of input time window, wavelet transform type and data balancing on the classification performance are considered, and a cost-sensitive algorithm is introduced to more accurately distinguish between normal and abnormal breathing events. PhysioNet apnea ECG database was used for training and evaluation. The 3-min Frequency B-Spline wavelets transform of ECG signal was carried out, and Dice Loss was used to train the classification model of sleep breathing. The classification accuracy was 93.44%, sensitivity was 88.9%, specificity was 96.2% and most indexes were better than other related work.

Semi-Supervised Learning for Low-cost Personalized Obstructive Sleep Apnea Detection Using Unsupervised Deep Learning and Single-Lead Electrocardiogram.

Obstructive sleep apnea (OSA) is a common sleep-related breathing disorder that can lead to a wide range of health issues if left untreated. This study aims to address the lack of research on personalized models for single-lead electrocardiogram (ECG)-based OSA detection, by proposing an automatic semi-supervised algorithm for automated low-cost personalization fine-tuning. We utilize a convolutional neural network (CNN)-based auto-encoder (AE) with a modified training objective to detect anomalous region of OSA. An indicator based on model outputs is utilized as a benchmark measure to assign pseudo-labels with confidence to each sample. Finally, we perform validation of the semi-supervised algorithm on the same database and cross-database scenarios. By introducing semi-supervised personalization, the accuracy, AUC, and mean absolute error (MAE) of the general model (GM) of 35 subjects from the same database are improved from 86.3%, 0.915, and 5.178 to 90.3%, 0.948, and 2.593. Simultaneously, in the validation of 25 subjects from a cross-database, the accuracy, AUC, and MAE of the GM are enhanced from 75.6%, 0.800, and 9.149 to 84.3%, 0.881, and 3.509. The improved version of AE demonstrates excellent adaptability in identifying abnormal features in OSA, employing a data-driven approach to assign pseudo-labels for unknown data automatically. Additionally, leveraging the pseudo-labels through a semi-supervised fine-tuning strategy provides a solution to overcome the limitation of clinical annotations, facilitating low-cost implementation of personalized models. The semi-supervised approach proposed in this paper provides a high-performance and annotation-free solution for personalized adjustment of automatic OSA detection.

P062 Validation of Unattended Polysomnography using Somte for Diagnosis of OSA in Pregnant Women during Early Gestation.

Abstract Introduction The detection of obstructive sleep apnoea (OSA) in pregnant women in early gestation using polysomnography (PSG) was logistically difficult during covid-19 lockdowns. Reliable alternates to PSG haven’t been well validated in early pregnancy. We compared the agreement between self-applied, unattended Somte and attended PSG, in pregnant women by 24 weeks gestation. Methods Women were recruited to undergo self-applied, unattended Somte and attended Grael PSG, by 24 weeks gestation. Tests were scored by a single experienced scorer (blinded to PSG result). An apnoea-hypopnea index (AHI) 5 or above at PSG was considered diagnostic of OSA. Diagnostic test analysis was undertaken (sensitivity, specificity, positive and negative predictive value (PPV/NPV)), by creating receiver operating characteristic (ROC) curves. The calculated AHI was compared and Bland-Altman plot used to plot agreement. Results Twenty-Four participants were included for analysis, median age 33.5(29.0, 37.0) years, gestation at PSG 14.1(13.4, 15.7) weeks and median BMI 31.2(26.3, 36.5). Five (20.8%) women had OSA. Median AHI were similar, 1.75 (0.50, 4.53) and 1.6 (0.58, 4.25) for Somte and PSG respectively. Somte demonstrated reasonable diagnostic test accuracy (area under the ROC curve 0.94 (95%CI 0.81-1.0)), sensitivity 80% (cut off AHI 5) and specificity 89.5%. The PPV and NPV were 80.0% and 94.7% respectively. Bland-Altman plot upper and lower confidence intervals of 6.37, -8.89, demonstrated lower and higher AHI scores slightly underestimated and overestimated, respectively. Discussion Somte provided reasonable accuracy compared to PSG in identifying OSA in early gestation in this sample. Larger studies will be helpful to confirm this finding.