Acoustic Analysis of Snoring in the Diagnosis of Obstructive Sleep Apnea Syndrome: A Call for More Rigorous Studies

Automatically detecting OSAHS patients based on transfer learning and model fusion

Objective. Snoring is the most typical symptom of obstructive sleep apnea hypopnea syndrome (OSAHS) that can be used to develop a non-invasive approach for automatically detecting OSAHS patients.Approach. In this work, a model based on transfer learning and model fusion was applied to classify simple snorers and OSAHS patients. Three kinds of basic models were constructed based on pretrained Visual Geometry Group-16 (VGG16), pretrained audio neural networks (PANN), and Mel-frequency cepstral coefficient (MFCC). The XGBoost was used to select features based on feature importance, the majority voting strategy was applied to fuse these basic models and leave-one-subject-out cross validation was used to evaluate the proposed model.Main results. The results show that the fused model embedded with top-5 VGG16 features, top-5 PANN features, and MFCC feature can correctly identify OSAHS patients (AHI > 5) with 100% accuracy.Significance. The proposed fused model provides a good classification performance with lower computational cost and higher robustness that makes detecting OSAHS patients at home possible.

Alternatives to Polysomnography for the Diagnosis of Pediatric Obstructive Sleep Apnea

Diagnosis of obstructive sleep apnea (OSA) in children with sleep-disordered breathing (SDB) requires hospital-based, overnight level I polysomnography (PSG). Obtaining a level I PSG can be challenging for children and their caregivers due to the costs, barriers to access, and associated discomfort. Less burdensome methods that approximate pediatric PSG data are needed. The goal of this review is to evaluate and discuss alternatives for evaluating pediatric SDB. To date, wearable devices, single-channel recordings, and home-based PSG have not been validated as suitable replacements for PSG. However, they may play a role in risk stratification or as screening tools for pediatric OSA. Further studies are needed to determine if the combined use of these metrics could predict OSA.

Snoring Index and Neck Circumference as Predictors of Adult Obstructive Sleep Apnea

Background. Snoring is the cardinal symptom of obstructive sleep apnea (OSA). The acoustic features of snoring sounds include intra-snore (including snoring index [SI]) and inter-snore features. However, the correlation between snoring sounds and the severity of OSA according to the apnea−hypopnea index (AHI) is still unclear. We aimed to use the snoring index (SI) and the Epworth Sleepiness Scale (ESS) to predict OSA and its severity according to the AHI among middle-aged participants referred for polysomnography (PSG). Methods. In total, 50 participants (mean age, 47.5 ± 12.6 years; BMI: 29.2 ± 5.6 kg/m2) who reported snoring and were referred for a diagnosis of OSA and who underwent a whole night of PSG were recruited. Results. The mean AHI was 30.2 ± 27.2, and the mean SI was 87.9 ± 56.3 events/hour. Overall, 11 participants had daytime sleepiness (ESS > 10). The correlation between SI and AHI (r = 0.33, p = 0.021) was significant. Univariate linear regression analysis showed that male gender, body mass index, neck circumference, ESS, and SI were associated with AHI. SI (β = 0.18, p = 0.004) and neck circumference (β = 2.40, p < 0.001) remained significantly associated with AHI by the multivariate linear regression model. Conclusion. The total number of snores per hour of sleep and neck circumference were positively associated with OSA among adults referred for PSG.

ResNet for recognition of Qi-deficiency constitution and balanced constitution based on voice

Background

According to traditional Chinese medicine theory, a Qi-deficiency constitution is characterized by a lower voice frequency, shortness of breath, reluctance to speak, an introverted personality, emotional instability, and timidity. People with Qi-deficiency constitution are prone to repeated colds and have a higher probability of chronic diseases and depression. However, a person with a Balanced constitution is relatively healthy in all physical and psychological aspects. At present, the determination of whether one has a Qi-deficiency constitution or a Balanced constitution are mostly based on a scale, which is easily affected by subjective factors. As an objective method of diagnosis, the human voice is worthy of research. Therefore, the purpose of this study is to improve the objectivity of determining Qi-deficiency constitution and Balanced constitution through one's voice and to explore the feasibility of deep learning in TCM constitution recognition.

Methods

The voices of 48 subjects were collected, and the constitution classification results were obtained from the classification and determination of TCM constitutions. Then, the constitution was classified according to the ResNet residual neural network model.

Results

A total of 720 voice data points were collected from 48 subjects. The classification accuracy rate of the Qi-deficiency constitution and Balanced constitution was 81.5% according to ResNet. The loss values of the model training and test sets gradually decreased to 0, while the ACC values of the training and test sets tended to increase, and the ACC values of the training set approached 1. The ROC curve shows an AUC value of 0.85.

Conclusion

The Qi-deficiency constitution and Balanced constitution determination method based on the ResNet residual neural network model proposed in this study can improve the efficiency of constitution recognition and provide decision support for clinical practice.

NaCl Ionization-Based Moisture Sensor Prepared by Aerosol Deposition for Monitoring Respiratory Patterns

A highly polarizable moisture sensor with multimodal sensing capabilities has great advantages for healthcare applications such as human respiration monitoring. We introduce an ionically polarizable moisture sensor based on NaCl/BaTiO₃ composite films fabricated using a facile aerosol deposition (AD) process. The proposed sensing model operates based on an enormous NaCl ionization effect in addition to natural moisture polarization, whereas all previous sensors are based only on the latter. We obtained an optimal sensing performance in a 0.5 µm-thick layer containing NaCl-37.5 wt% by manipulating the sensing layer thickness and weight fraction of NaCl. The NaCl/BaTiO₃ sensing layer exhibits outstanding sensitivity over a wide humidity range and a fast response/recovery time of 2/2 s; these results were obtained by performing the one-step AD process at room temperature without using any auxiliary methods. Further, we present a human respiration monitoring system using a sensing device that provides favorable and stable electrical signals under diverse respiratory scenarios.

Convolutional Neural Networks for Apnea Detection from Smartphone Audio Signals: Effect of Window Size

Although sleep apnea is one of the most prevalent sleep disorders, most patients remain undiagnosed and untreated. The gold standard for sleep apnea diagnosis, polysomnography, has important limitations such as its high cost and complexity. This leads to a growing need for novel cost-effective systems. Mobile health tools and deep learning algorithms are nowadays being proposed as innovative solutions for automatic apnea detection. In this work, a convolutional neural network (CNN) is trained for the identification of apnea events from the spectrograms of audio signals recorded with a smartphone. A systematic comparison of the effect of different window sizes on the model performance is provided. According to the results, the best models are obtained with 60 s windows (sensitivity-0.72, specilicity-0.89, AUROC = 0.88), For smaller windows, the model performance can be negatively impacted, because the windows become shorter than most apnea events, by which sound reductions can no longer be appreciated. On the other hand, longer windows tend to include multiple or mixed events, that will confound the model. This careful trade-off demonstrates the importance of selecting a proper window size to obtain models with adequate predictive power. This paper shows that CNNs applied to smartphone audio signals can facilitate sleep apnea detection in a realistic setting and is a first step towards an automated method to assist sleep technicians. Clinical Relevance- The results show the effect of the window size on the predictive power of CNNs for apnea detection. Furthermore, the potential of smartphones, audio signals, and deep neural networks for automatic sleep apnea screening is demonstrated.

Evaluating Prediction Models of Sleep Apnea From Smartphone-Recorded Sleep Breathing Sounds

Importance

Breathing sounds during sleep are an important characteristic feature of obstructive sleep apnea (OSA) and have been regarded as a potential biomarker. Breathing sounds during sleep can be easily recorded using a microphone, which is found in most smartphone devices. Therefore, it may be easy to implement an evaluation tool for prescreening purposes.

Objective

To evaluate OSA prediction models using smartphone-recorded sounds and identify optimal settings with regard to noise processing and sound feature selection.

Design, Setting, and Participants

A cross-sectional study was performed among patients who visited the sleep center of Seoul National University Bundang Hospital for snoring or sleep apnea from August 2015 to August 2019. Audio recordings during sleep were performed using a smartphone during routine, full-night, in-laboratory polysomnography. Using a random forest algorithm, binary classifications were separately conducted for 3 different threshold criteria according to an apnea hypopnea index (AHI) threshold of 5, 15, or 30 events/h. Four regression models were created according to noise reduction and feature selection from the input sound to predict actual AHI: (1) noise reduction without feature selection, (2) noise reduction with feature selection, (3) neither noise reduction nor feature selection, and (4) feature selection without noise reduction. Clinical and polysomnographic parameters that may have been associated with errors were assessed. Data were analyzed from September 2019 to September 2020.

Main Outcomes and Measures

Accuracy of OSA prediction models.

Results

A total of 423 patients (mean [SD] age, 48.1 [12.8] years; 356 [84.1%] male) were analyzed. Data were split into training (n = 256 [60.5%]) and test data sets (n = 167 [39.5%]). Accuracies were 88.2%, 82.3%, and 81.7%, and the areas under curve were 0.90, 0.89, and 0.90 for an AHI threshold of 5, 15, and 30 events/h, respectively. In the regression analysis, using recorded sounds that had not been denoised and had only selected attributes resulted in the highest correlation coefficient (r = 0.78; 95% CI, 0.69-0.88). The AHI (β = 0.33; 95% CI, 0.24-0.42) and sleep efficiency (β = -0.20; 95% CI, -0.35 to -0.05) were found to be associated with estimation error.

Conclusions and Relevance

In this cross-sectional study, recorded sleep breathing sounds using a smartphone were used to create reasonably accurate OSA prediction models. Future research should focus on real-life recordings using various smartphone devices.

Obstructive Sleep Apnea Detection Based on Sleep Sounds via Deep Learning

PURPOSE

This study aimed to propose a novel deep-learning method for automatic sleep apneic event detection and thus to estimate the apnea hypopnea index (AHI) and identify obstructive sleep apnea (OSA) in an event-by-event manner solely based on sleep sounds obtained by a noncontact audio recorder.

METHODS

We conducted a cross-sectional study of participants with habitual snoring or heavy breathing sounds during sleep to train and test a deep convolutional neural network named OSAnet for the detection of OSA based on sleep sounds. Polysomnography (PSG) was conducted, and sleep sounds were recorded simultaneously in a regular room without noise attenuation. The study was conducted in two phases. In phase one, eligible participants were enrolled and randomly allocated into training and validation groups for deep learning algorithm development. In phase two, eligible patients were enrolled in a test group for algorithm assessment. Sensitivity, specificity, accuracy, unweighted Cohen kappa coefficient (κ) and the area under the curve (AUC) were calculated using PSG as the reference standard.

RESULTS

A total of 135 participants were randomly divided into a training group (n, 116) and a validation group (n, 19). An independent test group of 59 participants was subsequently enrolled. Our algorithm achieved a precision of 0.81 and sensitivity of 0.78 in the test group for overall sleep event detection. The algorithm exhibited robust diagnostic performance to identify severe cases with a sensitivity of 95.6% and specificity of 91.6%.

CONCLUSION

Our results showed that a deep learning algorithm based on sleep sounds recorded by a noncontact voice recorder served as a feasible tool for apneic event detection and OSA identification. This technique may hold promise for OSA assessment in the community in a relatively comfortable and low-cost manner. Further studies to develop a tool based on a home-based setting are warranted.

Screening Severe Obstructive Sleep Apnea in Children with Snoring

Efficient screening for severe obstructive sleep apnea (OSA) is important for children with snoring before time-consuming standard polysomnography. This retrospective cross-sectional study aimed to compare clinical variables, home snoring sound analysis, and home sleep pulse oximetry on their predictive performance in screening severe OSA among children who habitually snored. Study 1 included 9 (23%) girls and 30 (77%) boys (median age of 9 years). Using univariate logistic regression models, 3% oxygen desaturation index (ODI3) ≥ 6.0 events/h, adenoidal-nasopharyngeal ratio (ANR) ≥ 0.78, tonsil size = 4, and snoring sound energy of 801–1000 Hz ≥ 22.0 dB significantly predicted severe OSA in descending order of odds ratio. Multivariate analysis showed that ODI3 ≥ 6.0 events/h independently predicted severe pediatric OSA. Among several predictive models, the combination of ODI3, tonsil size, and ANR more optimally screened for severe OSA with a sensitivity of 91% and a specificity of 94%. In Study 2 (27 (27%) girls and 73 (73%) boys; median age, 7 years), this model was externally validated to predict severe OSA with an accuracy of 76%. Our results suggested that home sleep pulse oximetry, combined with ANR, can screen for severe OSA more optimally than ANR and tonsil size among children with snoring.

Self-reported snoring is associated with chronic kidney disease in obese but not in normal-weight Chinese adults

Background

The relationship between sleeping disorders and chronic kidney disease (CKD) has already been reported. Snoring, a common clinical manifestation of obstructive sleep apnea–hypopnea syndrome, is of clinical value in assessing sleeping disorder severity. However, investigations of the connection between snoring and CKD are limited, especially in normal-weight populations. This study assessed the relationship between snoring frequency and CKD in obese and normal-weight people in China.

Methods

A community-based retrospective cross-sectional study of 3250 participants was performed. Study participants were divided into three groups – the regularly snoring group, occasionally snoring group, and never snoring group – based on their self-reported snoring frequency. CKD was defined as an estimated glomerular filtration rate of less than 60 mL/min/1.73 m². Multiple logistic regression analysis was used to explore the relevance between snoring frequency and CKD prevalence.

Results

The CKD prevalence in obese participants was higher than that in normal-weight participants. Frequent snorers had a higher prevalence of CKD than those who were not frequent snorers in the obese group. Snoring frequency was correlated with CKD prevalence in obese participants independent of age, sex, smoking and drinking status, systolic blood pressure, triglyceride level, high-density lipoprotein, and homeostasis model assessment of insulin resistance (odds ratio: 2.66; 95% CI: 1.36–5.19; p=.004), while the same relationships did not exist in normal-weight participants (odds ratio: 0.79; 95% CI: 0.32–1.98; p=.614).

Conclusions

Snoring appears to be independently associated with CKD in obese but not in normal-weight Chinese adults.

Perceptual snoring as a basis for a psychoacoustical modeling and clinical patient profiling

PURPOSE

The perceptual burden and social nuisance for mainly the co-sleeper can affect the relationship between snorer and bedpartner. Mandibular advancement devices (MAD) are commonly recommended to treat sleep-related breathing such as snoring or sleep apnea. There is no consensus about the definition of snoring particularly with MAD, which is essential for assessing the effectiveness of treatment. We aimed to stablish a notion of perceptual snoring with MAD in place.

METHODS

Sound samples, each 30 min long, were recorded during in-home, overnight, automatic mandibular repositioning titration studies in a population of 29 patients with obstructive sleep apnea syndrome (OSAS) from a clinical trial carried out to validate the MATRx plus. Three unspecialized and calibrated raters identified sound events and classified them as noise, snore, or breathing as well as providing scores for classification certainty and annoyance. Data were analyzed with respect to expiration-inspiration, duration, annoyance, and classification certainty.

RESULTS

A Fleiss' kappa (>0.80) and correlation duration of events (>0.90) between raters were observed. Prevalence of all breath sounds: snore 55.6% (N = 6398), breathing sounds 31.7% (N = 3652), and noise 9.3% (N = 1072). Inspiration occurs in 88.3% of events, 96.8% contained at least on expiration phase. Snore and breath events had similar duration, respectively 2.58s (sd 1.43) and 2.41s (sd 1.22). Annoyance is lowest for breathing events (8.00 sd 0.98) and highest for snore events (4.90 sd 1.92) on a VAS from zero to ten.

CONCLUSION

Perceptual sound events can be a basis for analysis in a psychosocial context. Perceived snoring occurs during both expiration as well as inspiration. Substantial amount of snoring remains despite repositioning of the mandible aimed at the reduction of AHI-ODI.

Snoring increases the development of coronary artery disease: a systematic review with meta-analysis of observational studies

PURPOSE

Snoring is one of the cardinal presentations of obstructive sleep apnea (OSA) and is more common than OSA. Abundant evidence has suggested a robust association between OSA and coronary artery disease (CAD). However, whether or not snoring alone is related to a higher risk of CAD is unknown. This study systematically reviewed observational studies with meta-analysis to evaluate the linkage between snoring and CAD.

METHODS AND RESULTS

We searched PubMed and Embase and retrieved 13 articles focusing on the relationship between snoring and CAD. These articles included a total of 151,366 participants and 9099 CAD patients. Quantitative analysis indicated that snoring was associated with a 28% (RR: 1.28, 95% CI: 1.13 to 1.45, P < 0.001) increase in the risk of developing CAD.

CONCLUSIONS

Snorers are exposed to a 28% increased risk for CAD. Although the association may be partly mediated through OSA, most snorers are not affected by apnea. Given the high prevalence of snoring and the disease burden of CAD in the general population, screening for snoring may be worthwhile for the early prevention of CAD.

Automatic Detection of Obstructive Sleep Apnea Events Using a Deep CNN-LSTM Model

Obstructive sleep apnea (OSA) is a common sleep-related respiratory disorder. Around the world, more and more people are suffering from OSA. Because of the limitation of monitor equipment, many people with OSA remain undetected. Therefore, we propose a sleep-monitoring model based on single-channel electrocardiogram using a convolutional neural network (CNN), which can be used in portable OSA monitor devices. To learn different scale features, the first convolution layer comprises three types of filters. The long short-term memory (LSTM) is used to learn the long-term dependencies such as the OSA transition rules. The softmax function is connected to the final fully connected layer to obtain the final decision. To detect a complete OSA event, the raw ECG signals are segmented by a 10 s overlapping sliding window. The proposed model is trained with the segmented raw signals and is subsequently tested to evaluate its event detection performance. According to experiment analysis, the proposed model exhibits Cohen's kappa coefficient of 0.92, a sensitivity of 96.1%, a specificity of 96.2%, and an accuracy of 96.1% with respect to the Apnea-ECG dataset. The proposed model is significantly higher than the results from the baseline method. The results prove that our approach could be a useful tool for detecting OSA on the basis of a single-lead ECG.

Audio-based snore detection using deep neural networks

BACKGROUND AND OBJECTIVE

Snoring is a prevalent phenomenon. It may be benign, but can also be a symptom of obstructive sleep apnea (OSA) a prevalent sleep disorder. Accurate detection of snoring may help with screening and diagnosis of OSA.

METHODS

We introduce a snore detection algorithm based on the combination of a convolutional neural network (CNN) and a recurrent neural network (RNN). We obtained audio recordings of 38 subjects referred to a clinical center for a sleep study. All subjects were recorded by a total of 5 microphones placed at strategic positions around the bed. The CNN was used to extract features from the sound spectrogram, while the RNN was used to process the sequential CNN output and to classify the audio events to snore and non-snore events. We also addressed the impact of microphone placement on the performance of the algorithm.

RESULTS

The algorithm achieved an accuracy of 95.3 ± 0.5%, a sensitivity of 92.2 ± 0.9%, and a specificity of 97.7 ± 0.4% over all microphones in snore detection on our data set including 18412 sound events. The best accuracy (95.9%) was observed from the microphone placed about 70 cm above the subject's head and the worst (94.4%) was observed from the microphone placed about 130 cm above the subject's head.

CONCLUSION

Our results suggest that our method detects snore events from audio recordings with high accuracy and that microphone placement does not have a major impact on detection performance.

Snoring patterns during home polysomnography. A proposal for a new classification

PURPOSE

Snoring is a very common disorder, but, at present, there is no universally accepted classification for the condition. The main aim of this paper is to introduce a home sleep monitoring-based classification of common snoring patterns in simple snorers and in patients with obstructive sleep apnea-hypopnea syndrome (OSAHS).

MATERIALS AND METHODS

In total, 561 consecutive patients with a history of snoring, either simple or associated with apnea, were enrolled in this home sleep monitoring study. Analysis of the polysomnographic traces and the snoring sensor allowed the main patterns of snoring and their characteristics to be determined.

RESULTS

Four patterns of snoring were identified. In a spectrum of increasing severity (mild, moderate or severe), snoring can be episodic, positional, continuous, or alternating, whereas in obstructive sleep apnea syndrome, the snoring events only occur between successive respiratory obstructive events. In mild snoring, the episodic pattern is the most frequent, whereas in moderate and severe snoring, the continuous snoring pattern occurs in most cases.

CONCLUSIONS

The proposed classification of snoring patterns would be beneficial for providing a realistic disturbance index, for the selection and evaluation of the outcomes of surgical techniques.

Self-reported snoring is associated with nonalcoholic fatty liver disease

Although nonalcoholic fatty liver disease (NAFLD) is associated with obstructive sleep apnea syndrome (OSAS), studies on the direct relationship between NAFLD and snoring, an early symptom of OSAS, are limited. We evaluated whether snorers had higher risk of developing NAFLD. The study was performed using data of the Tongmei study (cross-sectional survey, 2,153 adults) and Kailuan study (ongoing prospective cohort, 19,587 adults). In both studies, NAFLD was diagnosed using ultrasound; snoring frequency was determined at baseline and classified as none, occasional (1 or 2 times/week), or habitual (≥3 times/week). Odds ratios (ORs) and hazard ratios (HRs) with 95% confidence intervals were estimated using logistic and Cox models, respectively. During 10 years’ follow-up in Kailuan, 4,576 individuals with new-onset NAFLD were identified at least twice. After adjusting confounders including physical activity, perceived salt intake, body mass index (BMI), and metabolic syndrome (MetS), multivariate-adjusted ORs and HRs for NAFLD comparing habitual snorers to non-snorers were 1.72 (1.25–2.37) and 1.29 (1.16–1.43), respectively. These associations were greater among lean participants (BMI < 24) and similar across other subgroups (sex, age, MetS, hypertension). Snoring was independently and positively associated with higher prevalence and incidence of NAFLD, indicating that habitual snoring is a useful predictor of NAFLD, particularly in lean individuals.

Automatic Silence Events Detector from Smartphone Audio Signals: A Pilot mHealth System for Sleep Apnea Monitoring at Home

Obstructive sleep apnea (OSA) is a prevalent disease, but most patients remain undiagnosed and untreated. Recently, mHealth tools are being proposed to screen OSA patients at home. In this work, we analyzed full-night audio signals recorded with a smartphone microphone. Our objective was to develop an automatic detector to identify silence events (apneas or hypopneas) and compare its performance to a commercial portable system for OSA diagnosis (ApneaLink™, ResMed). To do that, we acquired signals from three subjects with both systems simultaneously. A sleep specialist marked the events on smartphone and ApneaLink signals. The automatic detector we developed, based on the sample entropy, identified silence events similarly than manual annotation. Compared to ApneaLink, it was very sensitive to apneas (detecting 86.2%) and presented an 83.4% positive predictive value, but it missed about half the hypopnea episodes. This suggests that during some hypopneas the flow reduction is not reflected in sound. Nevertheless, our detector accurately recognizes silence events, which can provide valuable respiratory information related to the disease. These preliminary results show that mHealth devices and simple microphones are promising non-invasive tools for personalized sleep disorders management at home.

Prediction of Apnea-Hypopnea Index Using Sound Data Collected by a Noncontact Device

OBJECTIVE

To predict the apnea-hypopnea index (AHI) in patients with obstructive sleep apnea (OSA) using data from breathing sounds recorded using a noncontact device during sleep.

STUDY DESIGN

Prospective cohort study.

SETTING

Tertiary referral hospital.

SUBJECT AND METHODS

Audio recordings during sleep were performed using an air-conduction microphone during polysomnography. Breathing sounds recorded from all sleep stages were analyzed. After noise reduction preprocessing, the audio data were segmented into 5-second windows and sound features were extracted. Estimation of AHI by regression analysis was performed using a Gaussian process, support vector machine, random forest, and simple linear regression, along with 10-fold cross-validation.

RESULTS

In total, 116 patients who underwent attended, in-laboratory, full-night polysomnography were included. Overall, random forest resulted in the highest performance with the highest correlation coefficient (0.83) and least mean absolute error (9.64 events/h) and root mean squared error (13.72 events/h). Other models resulted in somewhat lower but similar performances, with correlation coefficients ranging from 0.74 to 0.79. The estimated AHI tended to be underestimated as the severity of OSA increased. Regarding bias and precision, estimation performances in the severe OSA subgroup were the lowest, regardless of the model used. Among sound features, derivative of the area methods of moments of overall standard deviation demonstrated the highest correlation with AHI.

CONCLUSION

AHI was fairly predictable by using data from breathing sounds generated during sleep. The prediction model may be useful not only for prescreening but also for follow-up after treatment in patients with OSA.

Characterization of a tooth microphone coupled to an oral appliance device: A new system for monitoring OSA patients

Obstructive sleep apnea (OSA) is a highly prevalent chronic disease, especially in elderly and obese populations. Despite constituting a serious health, social and economic problem, most patients remain undiagnosed and untreated due to limitations in current equipment. In this work, we propose a novel method to diagnose OSA and monitor therapy adherence and effectiveness at home in a non-invasive and inexpensive way: combining acoustic analysis of breathing and snoring sounds with oral appliance therapy (OA). Audiodontics has introduced a new sensor, a tooth microphone coupled to an OA device, which is the main pillar of this system. The objective of this work is to characterize the response of this sensor, comparing it with a commercial tracheal microphone (Biopac transducer). Signals containing OSA-related sounds were acquired simultaneously with the two microphones for that purpose. They were processed and analyzed in time, frequency and time-frequency domains, in a custom MATLAB interface. We carried out a single-event approach focused on breaths, snores and apnea episodes. We found that the quality of the signals obtained by both microphones was quite similar, although the tooth microphone spectrum concentrated more energy at the high-frequency band. This opens a new field of study about high-frequency components of snores and breathing sounds. These characteristics, together with its intraoral position, wireless option and combination with customizable OAs, give the tooth microphone a great potential to reduce the impact of sleep disorders, by enabling prompt detection and continuous monitoring of patients at home.

Characterization of microphones for snoring and breathing events analysis in mHealth

Obstructive sleep apnea (OSA) is one of the most common sleep disorders, especially in elderly population. Despite its high prevalence and severe consequences, most patients remain undiagnosed due to serious limitations on the existing equipment. Efforts are being done to find cost-effective alternatives and mHealth solutions could play a key role. One promising approach in this context is the acoustic analysis of snoring. The sensor it requires is a microphone, which is widely available in different models and even integrated in smartphones. The objective of this work is to characterize and compare the responses of two commercial tracheal microphones and a mHealth-based microphone, as a proof-of-concept to evaluate their potential as sensors for OSA detection. To do that, we designed an experimental protocol to study OSA-related events (breaths, snores and apneas) simulated by 4 subjects. Test signals were simultaneously recorded with different microphones and posteriorly processed and analyzed. We accurately characterized the frequency response of the two commercial microphones, finding that one of them was too restrictive (bandwidth 50-250 Hz) and thus not suitable as snoring sensor for high-frequency acoustic analysis. Regarding smartphones, we studied the Samsung Galaxy S5 microphone. We found that, when located over the thorax, it provided quality signals comparable to those of tracheal microphones, with a broader frequency response. Further work is required, but this preliminary study suggests that acoustic analysis of snoring through mHealth solutions can be a feasible alternative to screen and monitor OSA patients at home.

Support vector machines for automated snoring detection: proof-of-concept

BACKGROUND

Snoring has been shown to be associated with adverse physical and mental health, independent of the effects of sleep disordered breathing. Despite increasing evidence for the risks of snoring, few studies on sleep and health include objective measures of snoring. One reason for this methodological limitation is the difficulty of quantifying snoring. Conventional methods may rely on manual scoring of snore events by trained human scorers, but this process is both time- and labor-intensive, making the measurement of objective snoring impractical for large or multi-night studies.

METHODS

The current study is a proof-of-concept to validate the use of support vector machines (SVM), a form of machine learning, for the automated scoring of an objective snoring signal. An SVM algorithm was trained and tested on a set of approximately 150,000 snoring and non-snoring data segments, and F-scores for SVM performance compared to visual scoring performance were calculated using the Wilcoxon signed rank test for paired data.

RESULTS

The ability of the SVM algorithm to discriminate snore from non-snore segments of data did not differ statistically from visual scorer performance (SVM F-score = 82.46 ± 7.93 versus average visual F-score = 88.35 ± 4.61, p = 0.2786), supporting SVM snore classification ability comparable to visual scorers.

CONCLUSION

In this proof-of-concept, we established that the SVM algorithm performs comparably to trained visual scorers, supporting the use of SVM for automated snoring detection in future studies.

Detecting Breathing and Snoring Episodes Using a Wireless Tracheal Sensor-A Feasibility Study

OBJECTIVE

Sleep-disordered breathing is both a clinical and a social problem. This implies the need for convenient solutions to simplify screening and diagnosis. The aim of the study was to investigate the sensitivity and specificity of a novel wireless system in detecting breathing and snoring episodes during sleep.

METHODS

A wireless acoustic sensor was elaborated and implemented. Segmentation (based on spectral thresholding and heuristics) and classification of all breathing episodes during recording were implemented through a mobile application. The system was evaluated on 1520 manually labeled episodes registered from 40 real-world, whole-night recordings of 16 generally healthy subjects.

RESULTS

The differentiation between normal breathing and snoring had 88.8% accuracy. As the system is intended for screening, high specificity of 95% is reported.

CONCLUSION

The system is a compromise between nonmedical phone applications and medical sleep studies. The presented approach enables the study to be repetitive, personal, and inexpensive. It has additional value in the form of well-recorded data which are reliable and comparable.

SIGNIFICANCE

The system opens unexplored possibilities in sleep monitoring and study enabling a multinight recording strategy involving the collection and analysis of abundant data from thousands of people.

Predicting Obstructive Sleep Apnea with Periodic Snoring Sound Recorded at Home

STUDY OBJECTIVES

The cost-effectiveness of diagnosing obstructive sleep apnea (OSA) could be improved by using a preliminary screening method among subjects with no suspicion of other sleep disorders. We aimed to evaluate the diagnostic value of periodic snoring sound recorded at home.

METHODS

We included 211 subjects, aged 18-83 (130 men), who were referred to our laboratory for suspicion of OSA, and had a technically successful overnight polygraphy, measured with the Nox T3 Sleep Monitor (Nox Medical, Iceland) with a built-in microphone. We analyzed the percentage of periodic snoring during the home sleep apnea study.

RESULTS

Apnea-hypopnea index (AHI) ranged from 0.1 to 116 events/h and the percentage of periodic snoring from 1% to 97%. We found a strong positive correlation (r = 0.727, p < 0.001) between periodic snoring and AHI. The correlation was slightly stronger among female, younger, and obese subjects. The best threshold value of the periodic snoring for predicting an AHI > 15 events/h with as high sensitivity as possible was found to be 15%. There, sensitivity was 93.3%, specificity 35.1%, and negative predictive value 75.0%.

CONCLUSIONS

According to our results, it is possible to set a periodic snoring threshold (15% or more) for the subject to advance to further sleep studies. Together with medical history and prior to more expensive studies, measuring periodic snoring at home is a simple and useful method for predicting the probability of OSA, in particular among women who are often unaware of their apnea-related snoring.

Continuous and Loud Snoring Only in the Supine Posture

Snoring and suspected sleep apneas are the most frequent causes for referral for a sleep study. Snoring varies across night and is usually recorded in all body postures. Here we report a unique patient showing continuous and loud snoring only in the supine posture.

Consumer sleep monitors: is there a baby in the bathwater?

The rapid expansion of consumer sleep devices is outpacing the validation data necessary to assess the potential use of these devices in clinical and research settings. Common sleep monitoring devices utilize a variety of sensors to track movement as well as cardiac and respiratory physiology. The variety of sensors and user-specific factors offer the potential, at least theoretically, for clinically relevant information. We describe the current challenges for interpretation of consumer sleep monitoring data, since the devices are mainly used in non-medical contexts (consumer use) although medically-definable sleep disorders may commonly occur in this setting. A framework for addressing questions of how certain devices might be useful is offered. We suggest that multistage validation efforts are crucially needed, from the level of sensor data and algorithm output, to extrapolations beyond healthy adults and into other populations and real-world environments.