INTRODUCTION: Obstructive sleep apnoea (OSA) is a serious but underdiagnosed condition. Demand for the gold standard diagnostic polysomnogram (PSG) far exceeds its availability. More efficient diagnostic methods are needed, even in tertiary settings. Machine learning (ML) models have strengths in disease prediction and early diagnosis. We explored the use of ML with oximetry, demographic and anthropometric data to diagnose OSA. METHODS: A total of 2,996 patients were included for modelling and divided into test and training sets. Seven commonly used supervised learning algorithms were trained with the data. Sensitivity (recall), specificity, positive predictive value (PPV) (precision), negative predictive value, area under the receiver operating characteristic curve (AUC) and F1 measure were reported for each model. RESULTS: In the best performing four-class model (neural network model predicting no, mild, moderate or severe OSA), a prediction of moderate and/or severe disease had a combined PPV of 94%; one out of 335 patients had no OSA and 19 had mild OSA. In the best performing two-class model (logistic regression model predicting no-mild vs. moderate-severe OSA), the PPV for moderate-severe OSA was 92%; two out of 350 patients had no OSA and 26 had mild OSA. CONCLUSION Our study showed that the prediction of moderate-severe OSA in a tertiary setting with an ML approach is a viable option to facilitate early identification of OSA. Prospective studies with home-based oximeters and analysis of other oximetry variables are the next steps towards formal implementation.
Humans ; Oximetry/methods* ; Sleep Apnea, Obstructive/diagnosis* ; Male ; Female ; Middle Aged ; Machine Learning ; Polysomnography ; Adult ; Anthropometry ; ROC Curve ; Aged ; Algorithms ; Predictive Value of Tests ; Sensitivity and Specificity ; Neural Networks, Computer ; Demography

Objective:To analyze the efficacy, influencing factors, and risk warning of multi-plane low-temperature plasma radiofrequency ablation（MLT-RFA） in the treatment of obstructive sleep apnea hypopnea syndrome（OSAHS）. Methods:A total of 118 OSAHS patients admitted from October 2022 to June 2024 were selected as the research subjects. They were divided into mild group（n=46）, moderate group（n=52）, and severe group（n=20） according to the severity of their condition. MLT-RFA treatment was used for all patients. After surgery, the results of polysomnography（PSG） and the changes in the Calier Sleep Apnea Quality of Life Index（SAQLI） were observed before and after treatment. The incidence of complications after treatment was recorded, and the clinical efficacy of the patients was evaluated. At the same time, they were divided into a treatment effective group（n=106） and an ineffective group（n=12） according to their effects. The general clinical data of the two groups were compared, and binary logistics regression analysis was conducted to identify independent factors that affect treatment efficacy and construct a model. ROC curve analysis was used to evaluate the diagnostic efficacy of the model. Results:The treatment effectiveness rate of the mild group was 93.48%, the moderate group was 90.38%, and the severe group was 80.00%. There was no statistically significant difference in the treatment effectiveness rate among the three groups（P>0.05）. The AHI of the mild group, moderate group, and severe group increased sequentially, while the LSaO2and SAQLI scores decreased sequentially. After treatment, the AHI of all three groups decreased compared to before treatment, while the LSaO2and SAQLI scores increased compared to before treatment, and the differences were statistically significant（P<0.05）. The pre-treatment AHI of the effective group was lower than that of the ineffective group, and the pre-treatment LSaO2and SAQLI were higher than those of the ineffective group, with statistically significant differences（P<0.05）. Pre-treatment LSaO2and pre-treatment SAQLI are independent factors affecting the efficacy of MLT-RFA（P<0.05）. The AUC of pre-treatment LSaO2, pre-treatment SAQLI, and combined prediction were 0.907, 0.763, and 0.947, respectively, with sensitivities of 0.896, 0.840, and 0.917, and specificities of 0.833, 0.667, and 0.887, respectively. Conclusion:MLT-RFA has a significant effect on the treatment of OSAHS, and the AHI, LSaO2, and SAQLI of patients before treatment can predict the treatment effect, with LSaO2 and SAQLI being independent influencing factors. The combinerd prediction model exhibits high diagnostic efficiency, sensitivity, and specificity.
Radiofrequency Ablation/methods* ; Plasma Gases ; Sleep Apnea, Obstructive/surgery* ; Polysomnography ; Postoperative Complications/epidemiology* ; Quality of Life ; Severity of Illness Index ; Treatment Outcome ; Humans

Recent studies have developed simple techniques for monitoring and assessing sleep. However, several issues remain to be solved for example high-cost sensor and algorithm as a home-use device. In this study, we aimed to develop an inexpensive and simple sleep monitoring system using a camera and video processing. Polysomnography (PSG) recordings were performed in six subjects for four consecutive nights. Subjects' body movements were simultaneously recorded by the web camera. Body movement was extracted by video processing from the video data and fi ve parameters were calculated for machine learning. Four sleep stages (WAKE, LIGHT, DEEP and REM) were estimated by applying these fi ve parameters to a support vector machine. The overall estimation accuracy was 70.3 ± 11.3% with the highest accuracy for DEEP (82.8 ± 4.7%) and the lowest for LIGHT (53.0 ± 4.0%) compared with correct sleep stages manually scored on PSG data by a sleep technician. Estimation accuracy for REM sleep was 68.0 ± 6.8%. The kappa was 0.19 ± 0.04 for all subjects. The present non-contact sleep monitoring system showed suffi cient accuracy in sleep stage estimation with REM sleep detection being accomplished. Low-cost computing power of this system can be advantageous for mobile application and modularization into home-device.
Machine Learning ; Methods ; Mobile Applications ; Polysomnography ; Sleep Stages ; Sleep, REM ; Support Vector Machine

BACKGROUND AND OBJECTIVES: The clinical significance and need for the treatment of primary snoring and mild obstructive sleep apnea have been recently questioned. In this study, we analyzed therapeutic outcome and the methods of treatment of such diseases. SUBJECTS AND METHOD: A retrospective review was conducted using the medical records of patients diagnosed with primary snoring or mild obstructive sleep apnea at a single institution from 2013 to 2015 through polysomnography or WATCHPAT. RESULTS: Of the 18 patients (37%) with primary snoring, 13 patients (72.2%) underwent surgery, four patients (22.2%) were treated with surgery and mandibular advancement device, and one patient (5.6%) underwent automatic positive airway pressure therapy. Of the 78 patients (61%) with mild obstructive sleep apnea, 35 patients (44.8%) had surgery, 24 patients (30.8%) were treated with mandibular advancement device, 13 patients (16.7%) were treated with surgery and mandibular advancement device and 6 patients (7.7%) received automatic positive airway pressure therapy. For primary snoring, while Epworth Sleepiness Scale and Pittsburg Sleep Quality Index did not improve, the snoring visual analog scale decreased significantly. In patients with mild obstructive sleep apnea, Apnea-Hypopnea Index, snoring decibel, Epworth Sleepiness Scale, and Pittsburg Sleep Quality Index were significantly decreased after treatment and the lowest oxygen saturation was significantly increased after treatment. CONCLUSION: For primary snoring, the direction of treatment should be determined in accordance with the presence of associated diseases related to sleep disturbance breathing. For mild obstructive sleep apnea, active treatment may be helpful.
Humans ; Mandibular Advancement ; Medical Records ; Methods ; Oxygen ; Polysomnography ; Respiration ; Retrospective Studies ; Sleep Apnea, Obstructive ; Snoring ; Visual Analog Scale

BACKGROUND AND OBJECTIVES: Obstructive sleep apnea (OSA) is highly prevalent in commercial vehicle operators (CMVOs). This study aimed to evaluate the poor sleep quality, daytime sleepiness, and the prevalence of self-reported OSA in CMVOs. SUBJECTS AND METHOD: We performed a retrospective review of the medical records of patients who visited a single institution with sleep problems from 2011 January to 2016 December. Among the patients, a total of 38 CMVOs was analyzed. Clinical information, questionnaires about sleep quality (Pittsburg sleep questionnaire, PSQI), excessive daytime sleepiness (Epworth sleepiness scale, ESS) and risk factors for OSA (STOP-Bang) were analyzed. The frequency of motor vehicle accidents and near accidents was assessed, and polysomnography (PSG) was used for OSA diagnosis purposes. RESULTS: The mean age of the study population was 45.3±11.8 years. The average score of PSQI, ESS, and STOP-Bang were 6.75±4.22, 10.79±7.12, and 4.62±3.34, respectively. A significant association between near accidents and high-risk group of OSA was observed [odds ratio (OR)=2.73, 95% confidence interval (CI)=1.08–4.48]. Subjects with poor sleep quality showed significantly increased risk of near accidents (OR=2.34, 95% CI=1.01–3.56). Receiver operating characteristic curves of STOP-Bang questionnaire using apnea-hypopnea index (cut-off value=5) indicates that suspected OSA group predicted by STOP-Bang score was significantly correlated with OSA severity (area under curve=0.72, sensitivity 77.1%, specificity 59.4%). CONCLUSION: Administration of STOP-Bang questionnaire before a PSG can identify high-risk subjects, supporting its further use in OSA screening of CMVOs.
Diagnosis ; Humans ; Korea ; Mass Screening ; Medical Records ; Methods ; Motor Vehicles ; Polysomnography ; Prevalence ; Retrospective Studies ; Risk Factors ; ROC Curve ; Sensitivity and Specificity ; Sleep Apnea, Obstructive ; Surveys and Questionnaires

OBJECTIVE: To evaluate the value of airway computed tomography (CT) in patients with obstructive sleep apnea (OSA) as a predictor of cerebrocardiovascular disease (CCVD) clinically, by quantitatively analyzing carotid arterial calcification (CarAC). MATERIALS AND METHODS: This study included 287 patients aged 40–80 years, who had undergone both polysomnography (PSG) and airway CT between March 2011 and October 2015. The carotid arterial calcium score (CarACS) was quantified using the modified Agatston method on each upper airway CT. The OSA severity was categorized as normal, mild, moderate, and severe using the PSG results. Clinical characteristics, comorbid diseases, and lipid profiles of all patients were analyzed, and the prevalence of CCVDs was investigated during the follow up period (52.2 ± 16.0 months). RESULTS: CCVD occurred in 27 patients (9.3%) at the end of follow-up, and the CCVD-present groups showed a significantly older mean age (57.5 years vs. 54.2 years), higher prevalence of hypertension (59% vs. 34%) and CarAC (51.9% vs. 20.8%), whereas sex, other comorbid diseases, and severity of OSA were not significantly different from the CCVD-absent group. A univariate analysis showed that age, hypertension, incidence of CarAC, and CarACS were risk factors for the occurrence of CCVD events. In a multivariate analysis, the incidence of CarAC was the only independent risk factor for CCVD. CONCLUSION: CarAC is an independent risk factor for CCVD, whereas the severity of OSA is not a contributory risk factor in patients with OSA. Therefore, additional analysis of CarACS based on airway CT scans may be useful for predicting CCVD.
Calcium ; Carotid Arteries ; Fluorouracil ; Follow-Up Studies ; Humans ; Hypertension ; Incidence ; Methods ; Multivariate Analysis ; Polysomnography ; Prevalence ; Risk Factors ; Sleep Apnea, Obstructive ; Tomography, X-Ray Computed

BACKGROUND AND OBJECTIVES: Upper airway obstruction can occur at the soft palate, tongue base, or epiglottis among obstructive sleep apnea (OSA) patients. Detection of these obstruction sites is very important for choosing a treatment modality for OSA. The purpose of this study was to evaluate the obstruction site of OSA patients and its association with mouth opening and head position. SUBJECTS AND METHOD: Forty-eight consecutive patients with suspicion of OSA were enrolled and underwent videofluoroscopy to evaluate the obstruction site, as well as polysomnography. Obstruction site, mouth opening, and head position were evaluated on videofluoroscopy, and their association was analyzed. RESULTS: According to the videofluoroscopy, 47 (97.9%) of 48 patients showed an obstruction in the soft palate, while 24 (50.0%) were located in the tongue base and 14 (29.2%) in the epiglottis. Multiple obstructions were observed in many patients. Mean apnea-hypopnea index was higher among patients with tongue base obstruction (42.3±26.7) compared to those without obstruction (26.4±21.2, p=0.058). However, epiglottis obstruction did not influence apnea-hypopnea index. Mouth opening did not show any association with tongue base obstruction (p=0.564), while head flexion was highly associated (p<0.001). CONCLUSION: Half of patients with OSA have tongue base obstruction, which worsens the apnea-hypopnea index. Head flexion is associated with tongue base obstruction, while mouth opening is not.
Airway Obstruction ; Epiglottis ; Head ; Humans ; Methods ; Mouth ; Palate, Soft ; Polysomnography ; Sleep Apnea, Obstructive ; Tongue

PURPOSE: Obstructive sleep apnea syndrome (OSAS) in young children is frequently caused by adenoid and/or tonsillar hypertrophy. Adenoidectomy is the first operative method for childhood chronic rhinosinusitis (CRS). We investigated factors associated with recurrent rhinosinusitis in preschool aged children with OSAS to determine the association of 2 common diseases. METHODS: One hundred forty-six children aged 2–5 years who were diagnosed as having OSAS after polysomnography between December 2003 and April 2016 were enrolled in this study. Children were divided into 2 groups with and without CRS. The 2 groups were compared in the severity of OSAS and allergy diseases and were evaluated for recurrent rhinosinusitis during the follow-up period, 1 year after diagnosis. RESULTS: Among 108 patients with OSAS who were followed up, 81 patients (75%) were diagnosed with CRS. There were no significant difference clinical and allergic characteristics between groups with and without CRS. However, bronchial asthma and otitis media was significantly more prevalent in patients with CRS than in those without (P=0.045 and P=0.000, respectively). Bronchial asthma and adenotonsillectomy was significantly associated with recurrent rhinosinusitis (P=0.005 and P=0.04, respectively) during the 1-year follow-up. CONCLUSION: Approximately 75% of preschool children with OSAS have suffered from CRS. Bronchial asthma is associated with CRS among OSAS children. Recurrent rhinosinusitis is decreased after adenotonsillectomy, and bronchial asthma is an associated factor for recurrent rhinosinusitis after a follow-up. This close relationship childhood OSAS and recurrent rhinosinusitis/bronchial asthma needs further studies to investigate their role in the association.
Adenoidectomy ; Adenoids ; Asthma ; Child ; Child, Preschool* ; Diagnosis ; Follow-Up Studies ; Humans ; Hypersensitivity ; Hypertrophy ; Methods ; Otitis Media ; Polysomnography ; Rhinitis ; Sinusitis ; Sleep Apnea, Obstructive*

The polysomnogram (PSG) analysis is considered the golden standard for sleep staging under the clinical environment. The electroencephalogram (EEG) signal is the most important signal for classification of sleep stages. However, in-vivo signal recording and analysis of EEG signal presents us with a few technical challenges. Electrocardiogram signals on the other hand, are easier to record, and can provide an attractive alternative for home sleep monitoring. In this paper we describe a method based on deep neural network (DNN), which can be used for the classification of the sleep stages into Wake (W), rapid-eye-movement (REM) and non-rapid-eye-movement (NREM) sleep stage. We apply the sleep stage stacked autoencoder to constitute a 4-layer DNN model. In order to test the accuracy of our method, eighteen PSGs from the MIT-BIH Polysomnographic Database were used. A total of 11 features were extracted from each electrocardiogram recording The experimental design employs cross-validation across subjects, ensuring the independence of the training and the test data. We obtained an accuracy of 77% and a Cohen's kappa coefficient of about 0.56 for the classification of Wake, REM and NREM.
Classification ; Electrocardiography ; Electroencephalography ; Hand ; Methods ; Polysomnography ; Research Design ; Sleep Stages

OBJECTIVES: Polysomnography is essential to diagnose sleep disorders. It is used to identify a patient's sleep pattern during sleep. This pattern is obtained by a doctor or health practitioner by using a scoring process, which is time consuming. To overcome this problem, we developed a system that can automatically classify sleep stages. METHODS: This paper proposes a new method for sleep stage classification, called the fast convolutional method. The proposed method was evaluated against two sleep datasets. The first dataset was obtained from physionet.org, a physiologic signals data centers. Twenty-five patients who had a sleep disorder participated in this data collection. The second dataset was collected in Mitra Keluarga Kemayoran Hospital, Indonesia. Data was recorded from ten healthy respondents. RESULTS: The proposed method reached 73.50% and 56.32% of the F-measures for the PhysioNet and Mitra Keluarga Kemayoran Hospital data, respectively. Both values were the highest among all the machine learning methods considered in this study. The proposed method also had an efficient running time. The fast convolutional models of the PhysioNet and Mitra Keluarga Kemayoran Hospital data needed 42.60 and 0.06 seconds, respectively. CONCLUSIONS: The fast convolutional method worked well on the tested datasets. It achieved a high F-measure result and an efficient running time. Thus, it can be considered a promising tool for sleep stage classification.
Classification* ; Data Collection ; Dataset ; Humans ; Indonesia ; Machine Learning ; Methods* ; Polysomnography ; Running ; Sleep Stages* ; Sleep Wake Disorders ; Surveys and Questionnaires