Snoring Signals Research Articles

Snoring Sound Recognition Using Multi-Channel Spectrograms

Obstructive sleep apnea-hypopnea syndrome (OSAHS) is a common and high-risk sleep-related breathing disorder. Snoring detection is a simple and non-invasive method. In many studies, the feature maps are obtained by applying a short-time Fourier transform (STFT) and feeding the model with single-channel input tensors. However, this approach may limit the potential of convolutional networks to learn diverse representations of snore signals. This paper proposes a snoring sound detection algorithm using a multi-channel spectrogram and convolutional neural network (CNN). The sleep recordings from 30 subjects at the hospital were collected, and four different feature maps were extracted from them as model input, including spectrogram, Mel-spectrogram, continuous wavelet transform (CWT), and multi-channel spectrogram composed of the three single-channel maps. Three methods of data set partitioning are used to evaluate the performance of feature maps. The proposed feature maps were compared through the training set and test set of independent subjects by using a CNN model. The results show that the accuracy of the multi-channel spectrogram reaches 94.18%, surpassing that of the Mel-spectrogram that exhibits the best performance among the single-channel spectrograms. This study optimizes the system in the feature extraction stage to adapt to the superior feature learning capability of the deep learning model, providing a more effective feature map for snoring detection.

A Novel Enhancement Approach Following MVMD and NMF Separation of Complex Snoring Signals.

Snoring is a prominent characteristic of sleep-disordered breathing, and its detection is critical for determining the severity of the upper airway obstruction and improving daily quality of life. Home snoring analysis is a highly invasive method, but it becomes challenging when a sleeping partner also snores, leading to distorted evaluations in such environments. In this article, we tackle the problem of complex snore signal separation of multiple snorers. This article introduces two audio-based methods that efficiently extract an individual's snoring signal, allowing for the analysis of sleep-breathing disorders in a normal sleeping environment without isolating individuals. In the first method, Principal Component Analysis (PCA) identifies the source components from the finite number of modes generated by the decomposition of the snoring mixture using Multivariate Variational Mode Decomposition (MVMD). The second method applies Blind Source Separation (BSS) based on Non-Negative Matrix Factorization (NMF) to separate the single-channel snoring mixture. Furthermore, the decomposed signals are tuned using the iterative enhancement algorithm to adequately match the source snoring signals. These methods were evaluated by simulating various real-time snoring recordings of 7 subjects (2 men, 2 women, and 3 children). The correlation coefficient between the source and its separated signal was computed to assess the separation results, exhibiting good performance of the methods used. The enhancement approach also demonstrated its efficiency by increasing the correlation over to 80% in both methods. The experimental results show that the proposed algorithms are effective and practical for separating mixed snoring signals.

Battling with the low-resource condition for snore sound recognition: introducing a meta-learning strategy

Snoring affects 57 % of men, 40 % of women, and 27 % of children in the USA. Besides, snoring is highly correlated with obstructive sleep apnoea (OSA), which is characterised by loud and frequent snoring. OSA is also closely associated with various life-threatening diseases such as sudden cardiac arrest and is regarded as a grave medical ailment. Preliminary studies have shown that in the USA, OSA affects over 34 % of men and 14 % of women. In recent years, polysomnography has increasingly been used to diagnose OSA. However, due to its drawbacks such as being time-consuming and costly, intelligent audio analysis of snoring has emerged as an alternative method. Considering the higher demand for identifying the excitation location of snoring in clinical practice, we utilised the Munich-Passau Snore Sound Corpus (MPSSC) snoring database which classifies the snoring excitation location into four categories. Nonetheless, the problem of small samples remains in the MPSSC database due to factors such as privacy concerns and difficulties in accurate labelling. In fact, accurately labelled medical data that can be used for machine learning is often scarce, especially for rare diseases. In view of this, Model-Agnostic Meta-Learning (MAML), a small sample method based on meta-learning, is used to classify snore signals with less resources in this work. The experimental results indicate that even when using only the ESC-50 dataset (non-snoring sound signals) as the data for meta-training, we are able to achieve an unweighted average recall of 60.2 % on the test dataset after fine-tuning on just 36 instances of snoring from the development part of the MPSSC dataset. While our results only exceed the baseline by 4.4 %, they still demonstrate that even with fine-tuning on a few instances of snoring, our model can outperform the baseline. This implies that the MAML algorithm can effectively tackle the low-resource problem even with limited data resources.

Wearable Sleep Monitoring System Based on Machine Learning Using Snoring Sound Signal

Abstract According to the obstructive sleep apnea Syndrome (OSAS), a wearable sleep monitoring system is designed based on machine learning using snoring sound signal. The system picks up snoring signal via bone conduction sensor, and calculates the apnea-hypopnea index (AHI). By analyzing the snoring signal in frequency domain, spectral entropy and other frequency-domain features are selected. Finally, the neural network classifier model is established. In the model, the input variables are eight frequency-domain features, and the output response is related to AHI value. Trained by machine learning, the result shows that the average accuracy in identifying the severity of the four kinds of OSAS categories is 59%. The system uses the measured data of snoring to analyze the symptoms of OSAS, so as to realize the preliminary forecast based on the snoring data. The system proposed in this paper has a good application development prospect in intelligent monitoring and medical instruments.

Using entropy of snoring, respiratory effort and electrocardiography signals during sleep for OSA detection and severity classification

Study objectivesObstructive sleep apnea (OSA) is a very prevalent disease and its diagnosis is based on polysomnography (PSG). We investigated whether snoring-sound-, very low frequency electrocardiogram (ECG-VLF)- and thoraco-abdominal effort- PSG signal entropy values could be used as surrogate markers for detection of OSA and OSA severity classification. MethodsThe raw data of the snoring-, ECG- and abdominal and thoracic excursion signal recordings of two consecutive full-night PSGs of 86 consecutive patients (22 female, 53.74 ± 12.4 years) were analyzed retrospectively. Four epochs (30 s each, manually scored according to the American Academy of Sleep Medicine standard) of each sleep stage (N1, N2, N3, REM, awake) were used as the ground truth. Sampling entropy (SampEn) of all the above signals was calculated and group comparisons between the OSA severity groups were performed. In total, (86x4x5 = )1720 epochs/group/night were included in the training set as an input for a support vector machine (SVM) algorithm to classify the OSA severity classes. Analyses were performed for first- and second-night PSG recordings separately. ResultsTwenty-seven patients had mild (RDI = ≥ 5/h but <15/h), 21 patients moderate (RDI ≥15/h but <30/h) and 23 patients severe OSA (RDI ≥30/h). Fifteen patients had an RDI <5/h and were therefore considered non-OSA. Using SE on the above three PSG signal data and using a SVM pipeline, it was possible to distinguish between the four OSA severity classes. The best metric was snoring signal-SE. The area-under-the-curve (AUC) calculations showed reproducible significant results for both nights of PSG. The second night data were even more significant, with non-OSA (R) vs. light OSA (L) 0.61, R vs. moderate (M) 0.68, R vs. heavy OSA (H) 0.84, L vs. M 0.63, M vs. H 0.65 and L vs. H 0.82. The results were not confounded by age or gender. ConclusionsSampEn of either snoring-, very low ECG-frequencies- or thoraco-abdominal effort signals alone may be used as a surrogate marker to diagnose OSA and even predict OSA severity. More specifically, in this exploratory study snoring signal SampEn showed the greatest predictive accuracy for OSA among the three signals. Second night data showed even more accurate results for all three parameters than first-night recordings. Therefore, technologies using only parts of the PSG signal, e.g. sound-recording devices, may be used for OSA screening and OSA severity group classification.

Pilot study of contactless sleep apnea detection based on snore signals with hardware implementation

Objective. Sleep apnea has a high incidence and is a potentially dangerous disease, and its early detection and diagnosis are challenging. Polysomnography (PSG) is considered the best approach for sleep apnea detection, but it requires cumbersome and complicated operations. Thus, it cannot satisfy the family healthcare needs. Approach. To facilitate the initial detection of sleep apnea in the home environment, we developed a sleep apnea classification model based on snoring and hybrid neural network, and implemented the well trained model in an embedded hardware platform. We used snore signals from 32 patients at Shenzhen People's Hospital. The Mel-Fbank features were extracted from snore signals to build a sleep apnea classification model based on Bi-LSTM with attention mechanism. Main results. The proposed model classified snore signals into four types: hypopnea, normal condition, obstructive sleep apnea, and central sleep apnea, with 83.52% and 62.31% accuracies, corresponding to the subject-dependence and subject-independence validation, respectively. After pruning and model quantization, at the cost of 0.81% and 0.95% accuracy loss of the subject dependence and subject independence classification, respectively, the number of model parameters and model storage space were reduced by 32.12% and 60.37%, respectively. The model exhibited accuracies of 82.71% and 61.36% based on the subject dependence and subject independence validations, respectively. When the well trained model was successfully porting and running on an STM32 ARM-embedded platform, the model accuracy was 58.85% for the four classifications based on leave-one-subject-out validation. Significance. The proposed sleep apnea detection model can be used in home healthcare for the initial detection of sleep apnea.

Design of embedded real-time system for snoring and OSA detection based on machine learning

Obstructive sleep apnea (OSA) is a common sleep disorder. As a gold standard to detect OSA, the polysomnography (PSG) is widely used. However, the acquisition of nighttime PSG signals is very inconvenient. Compared to PSG, snoring-based OSA detection is low-cost, convenient and non-invasive. In this work, the apnea situation is studied based on the nighttime snoring signals to detect OSA. Firstly, a nighttime audio database regarding snoring sounds is built by the multi-threshold endpoint detection from original night recordings. Next, feature extraction is carried out on the segregated sound signals to obtain the characteristic information of snoring. Then, feature categories are predicted based on the machine learning models. The proposed algorithm is transplanted to the embedded device to build a real-time system which can detect snoring and OSA events. The study experiments with a variety of detection schemes and finally trains a multi-classification temporal convolutional network (TCN) to classify night audio as non-snoring, snoring or OSA-related snoring. The experiment results have achieved good performance that the detection accuracy of OSA-related snoring is up to 96.7 %. In order to further promote this method, the trained TCN model is transplanted to an embedded device to detect snoring and OSA events using as a real-time system. The experimental results fully show that the proposed detection scheme can achieve high accuracy for the detection of snoring and OSA events, which demonstrates that it can be used a new tool to conveniently help potential OSA patients understand their sleep health for home health care.

Automatically detecting apnea-hypopnea snoring signal based on VGG19 + LSTM

Snoring is a typical syndrome of obstructive sleep apnea hypopnea syndrome (OSAHS). The acoustic analysis of snoring sound has been proved potential to develop a non-invasive approach for assisting diagnose OSAHS. In this work, a pre-trained VGG19 and the long short-term memory (LSTM) fused model was proposed to classify snoring sounds of simple snorers and OSAHS patients and detect apnea-hypopnea snoring from the whole night recorded sounds of patients. Mel-spectrograms of snoring sounds were fed into the VGG19 + LSTM model to learn relatively distinguishable features. Compared with other fused models, the proposed VGG19 + LSTM model yielded the highest accuracy of 99.31 % in classifying simple snorers’ snoring and OSAHS patients’ snoring. For distinguishing normal snoring and apnea-hypopnea snoring of patients, the VGG19 + LSTM achieved 85.21 % and 66.29 % accuracies based on hold-out and leave-one-subject-out validation methods respectively. The estimated AHI highly correlated with PSG AHI with a Pearson correlation coefficient of 0.966 (p < 0.001). Results of the proposed model demonstrate that acoustic analysis of snoring sounds has great potential for screening sleep and diagnosing OSAHS.

Contactless sleep apnea detection in snoring signals using hybrid deep neural networks targeted for embedded hardware platform with real-time applications

It is estimated that 936 million adults aged 30–69 years suffer from mild to severe obstructive sleep apnea that can result in poor sleep quality and even endanger their lives. To facilitate the primary screening of sleep apnea in the home environment, a contactless and automated detection method of sleep apnea is proposed based on snoring signals and hybrid deep neural networks with real-time applications. In our work, 5700 snoring segments were collected from 32 patients. The Mel filter banks (FBanks) features were extracted from the snoring sounds and a classification model was constructed based on hybrid deep neural networks. By taking advantage of the pruning and quantization methods, the classification model was transplanted to an STM32 ARM embedded hardware platform with real-time application. Results showed that through the model pruning, with the tradeoff of losing 0.35% accuracy, the number of neural network parameters was reduced by 40.27%. When the pruned parameters were encapsulated, the speed of the convolution operation increased by 27.02%. In addition, the averaged classification accuracy of obstructive apnea snores, central apnea snore, hypopnea, and normal snores was 74.27%. The results demonstrate that our proposed contactless sleep apnea detection system, targeted for embedded hardware platform with real-time applications, is capable of identifying three kinds of apnea diseases and normal snores, which can play a major role in the primary screening and follow-up treatment of sleep apnea for in-home healthcare applications.

Comparison of Drug-Induced Sleep Endoscopy and Natural Sleep Endoscopy in the Assessment of Upper Airway Pathophysiology During Sleep: Protocol and Study Design.

Study Objectives: Obstructive sleep apnea (OSA) is increasingly recognized as a complex and heterogenous disorder. As a result, a “one-size-fits-all” management approach should be avoided. Therefore, evaluation of pathophysiological endotyping in OSA patients is emphasized, with upper airway collapse during sleep as one of the main features. To assess the site(s) and pattern(s) of upper airway collapse, natural sleep endoscopy (NSE) is defined as the gold standard. As NSE is labor-intensive and time-consuming, it is not feasible in routine practice. Instead, drug-induced sleep endoscopy (DISE) is the most frequently used technique and can be considered as the clinical standard. Flow shape and snoring analysis are non-invasive measurement techniques, yet are still evolving. Although DISE is used as the clinical alternative to assess upper airway collapse, associations between DISE and NSE observations, and associated flow and snoring signals, have not been quantified satisfactorily. In the current project we aim to compare upper airway collapse identified in patients with OSA using endoscopic techniques as well as flow shape analysis and analysis of tracheal snoring sounds between natural and drug-induced sleep.Methods: This study is a blinded prospective comparative multicenter cohort study. The study population will consist of adult patients with a recent diagnosis of OSA. Eligible patients will undergo a polysomnography (PSG) with NSE overnight and a DISE within 3 months. During DISE the upper airway is assessed under sedation by an experienced ear, nose, throat (ENT) surgeon using a flexible fiberoptic endoscope in the operating theater. In contrast to DISE, NSE is performed during natural sleep using a pediatric bronchoscope. During research DISE and NSE, the standard set-up is expanded with additional PSG measurements, including gold standard flow and analysis of tracheal snoring sounds.Conclusions: This project will be one of the first studies to formally compare collapse patterns during natural and drug-induced sleep. Moreover, this will be, to the authors' best knowledge, the first comparative research in airflow shape and tracheal snoring sounds analysis between DISE and NSE. These novel and non-invasive diagnostic methods studying upper airway mechanics during sleep will be simultaneously validated against DISE and NSE.Clinical Trial Registration: www.ClinicalTrials.gov, identifier: NCT04729478.

Unsupervised Approach for the Identification of the Predominant Site of Upper Airway Collapse in Obstructive Sleep Apnoea Patients Using Snore Signals.

Knowledge regarding the site-of-collapse in the upper airway in obstructive sleep apnoea (OSA) has implications for treatment options and their outcomes. However, current methods to identify the site-of-collapse are not suitable for clinical practice due to the invasive nature, the time/cost of the tests and the inconsistency of the obstruction site identified with natural and drug-induced sleep. In this study, we adopted an unsupervised algorithm to identify the predominant site-of-collapse of the upper airway during natural sleep using nocturnal audio recordings. Nocturnal audio was recorded together with full-night polysomnography using a ceiling microphone. Various acoustic features of the snore signal during hypopnoea events were extracted. We developed a feature selection algorithm combining silhouette analysis with the Laplacian score algorithm to select the high performing features. A k-means clustering model was developed to form clusters using the features extracted from snore data and analyse the correlation between the clusters generated and the predominant site-of-collapse. Cluster analysis showed that the data tends to fit well in two clusters with a mean silhouette coefficient of 0.79 and with an accuracy of 68% for classifying tongue/non-tongue collapse. The results indicate a correlation between snoring and the predominant site-of-collapse. Therefore, it could potentially be used as a practical, non-invasive, low-cost diagnosis tool for improving the selection of appropriate therapy for OSA patients without any additional burden to the patients undergoing a sleep test.

Association of snoring characteristics with predominant site of collapse of upper airway in obstructive sleep apnea patients.

Acoustic analysis of isolated events and snoring by previous researchers suggests a correlation between individual acoustic features and individual site of collapse events. In this study, we hypothesized that multiparameter evaluation of snore sounds during natural sleep would provide a robust prediction of the predominant site of airway collapse. The audio signals of 58 obstructive sleep apnea patients were recorded simultaneously with full-night polysomnography. The site of collapse was determined by manual analysis of the shape of the airflow signal during hypopnea events and corresponding audio signal segments containing snore were manually extracted and processed. Machine learning algorithms were developed to automatically annotate the site of collapse of each hypopnea event into three classes (lateral wall, palate, and tongue base). The predominant site of collapse for a sleep period was determined from the individual hypopnea annotations and compared to the manually determined annotations. This was a retrospective study that used cross-validation to estimate performance. Cluster analysis showed that the data fit well in two clusters with a mean silhouette coefficient of 0.79 and an accuracy of 68% for classifying tongue/non-tongue collapse. A classification model using linear discriminants achieved an overall accuracy of 81% for discriminating tongue/non-tongue predominant site of collapse and accuracy of 64% for all site of collapse classes. Our results reveal that the snore signal during hypopnea can provide information regarding the predominant site of collapse in the upper airway. Therefore, the audio signal recorded during sleep could potentially be used as a new tool in identifying the predominant site of collapse and consequently improving the treatment selection and outcome.

Real-time prediction of upcoming respiratory events via machine learning using snoring sound signal.

The aim of the study was to inspect the acoustic properties and sleep characteristics of a preapneic snoring sound. The feasibility of forecasting upcoming respiratory events by snoring sound was also investigated. Participants with habitual snoring or a heavy breathing sound during sleep were recruited consecutively. Polysomnography was conducted, and snoring-related breathing sound was recorded simultaneously. Acoustic features and sleep features were extracted from 30-second samples, and a machine learning algorithm was used to establish 2 prediction models. A total of 74 eligible participants were included. Model 1, tested by 5-fold cross-validation, achieved an accuracy of 0.92 and an area under the curve of 0.94 for respiratory event prediction. Model 2, with acoustic features and sleep information tested by Leave-One-Out cross-validation, had an accuracy of 0.78 and an area under the curve of 0.80. Sleep position was found to be the most important among all sleep features contributing to the performance of the 2 models. Preapneic sound presented unique acoustic characteristics, and snoring-related breathing sound could be deployed as a real-time apneic event predictor. The models, combined with sleep information, serve as a promising tool for an early warning system to forecast apneic events. Wang B, Yi X, Gao J, etal. Real-time prediction of upcoming respiratory events via machine learning using snoring sound signal. J Clin Sleep Med. 2021;17(9):1777-1784.

Generalized Subspace Snoring Signal Enhancement Based on Noise Covariance Matrix Estimation

Acoustical properties of snoring signal have been widely studied as a potentially cost-effective and reliable alternative to diagnosing obstructive sleep apnea hypopnea syndrome, with a common recognition that the diagnostic accuracy depends heavily on the snoring signal quality. In the paper, generalized subspace noise reduction based on noise covariance matrix estimate is proposed. The noise covariance matrix is the Toeplitz matrix of the unbiased autocorrelation sequence which is estimated by recursive averaging its past value adjusted by a time-varying smoothing parameter controlled by the snoring signal presence probability, and the signal presence is determined by the ratio of temporal frame autocorrelation value to its minimum absolute value. The proposed method has a better estimate of noise covariance matrix, and the results of objective quality measurements and spectrograms of snoring signal show obvious improvement in terms of noise reduction and signal distortion under different non-stationary noise environments compared with conventional subspace enhancement algorithm.

Polysomnographic characteristics and acoustic analysis of catathrenia (nocturnal groaning)

Objective: Catathrenia is a sleep disorder characterized by nocturnal groaning sounds emitted during prolonged expiration. As a rare condition, its polysomnographic findings were inconsistent. We aimed to present polysomnographic characteristics of catathrenia patients and perform acoustic analysis of groaning sounds. Approach: Twenty-three patients (eight males and 15 females) diagnosed with catathrenia by video-polysomnography were included. They underwent clinical evaluation and physical examination, and answered a questionnaire. Acoustic analyses (oscillograms and spectrograms) of catathrenia and snoring signals were performed by Praat 6.1.09. Sounds were classified according to Yanagihara criteria. Main results: The average age of catathrenia patients was 29.6 ± 10.0 years, with a body mass index of 22.3 ± 5.1 kg m−2. A total of 3728 groaning episodes were documented. Catathrenia events of 16 patients (70%) were rapid eye movement (REM)-predominant. The average duration of groaning was 11.4 ± 4.6 s, ranging from 1.3 to 74.9 s. All signals of groaning were rhythmic or semi-rhythmic, classified as type I and type II, respectively, with formants and harmonics. Snoring events were observed in nine patients. Snoring mainly occurred in the non-REM stage, with a duration of less than 1.5 s. Signals of snoring were chaotic, classified as type III, without harmonics. Significance: Catathrenia occurred in all sleep stages but mainly in REM. Durations of groaning varied greatly across patients. Acoustic characteristics of catathrenia were typical. Groaning had rhythmic or semi-rhythmic waveform, formants and harmonics, indicating vocal origin, while snoring had chaotic waveform.

Automated identification of the predominant site of upper airway collapse in obstructive sleep apnoea patients using snore signal

Objective: This study provides a novel approach for an automated system using a machine learning algorithm to predict the predominant site of upper airway collapse into four classes (‘lateral wall’, ‘palate’, ‘tongue-base’ related collapse or ‘multi-level’ site-of-collapse) in obstructive sleep apnoea (OSA) patients from the audio signal recorded during normal sleep. Approach: Snore sounds from 58 patients were recorded simultaneously with full-night polysomnography during sleep with a ceiling mounted microphone. The probable site-of-airway collapse was determined by manual analysis of the shape of the airflow signal during hypopnoea. Time and frequency features of the audio signal were extracted from each hypopnoea event to classify the audio signal into ‘lateral wall’, ‘palate’ and ‘tongue-base’ related collapse according to prior research. The data was divided into two sets. The Learning Set contained the data of the first 45 patients and was used for building the model. The Hidden Set contained the data from the remaining 13 patients and was used for testing the performance of the model. Feature selection was employed to boost the classification performance. The classification was carried out with a multi-class linear discriminant analysis classifier to classify the predominant site-of-collapse for a patient into the four classes. Performance was evaluated by comparing the automatic and manually labelled data based on the predominant site-of-collapse and calculating the accuracy. Main results: The model achieved an overall accuracy on the Hidden Set of 77% for discriminating tongue/non-tongue collapse and an accuracy of 62% accuracy for all site-of-collapse classes. Significance: Our results demonstrate that the audio signal recorded during sleep can successfully identify the site-of-collapse in the upper airway. The additional information regarding the obstruction site may assist clinicians in deciding the most appropriate treatment for OSA.

Identifying the Predominant Site of Upper Airway Collapse in Obstructive Sleep Apnoea Patients Using Snore Signals.

Knowledge regarding the site of airway collapse could help in choosing an appropriate structure-specific or individualized treatment for obstructive sleep apnoea (OSA). We investigated if the audio signal recorded during hypopnoea (partial obstruction) events can predict the site-of-collapse of the upper airway. In this study, we designed an automatic classifier that predicts the predominant site of upper airway collapse for a patient as "lateral wall", "palate", "tongue-based" related collapse or "multi-level" site-of-collapse by processing of the audio signal. The probable site-of-collapse was determined by manual analysis of the shape of the airflow signal during hypopnoea, which has been reported to correlate with the site of collapse. Audio signal was recorded simultaneously with full-night polysomnography during sleep with a ceiling microphone. Various time and frequency features of the audio signal were extracted to classify the audio signal into lateral wall, palate and tongue-base related collapse. We introduced an unbiased process using nested leave-one patient-out cross-validation to choose the optimal features. The classification was carried out with a multi-class linear discriminant analysis classifier. Performance of the proposed model showed that our automatic system can achieve an overall accuracy of 65% for determining the predominant site-of-collapse for all site-of-collapse classes and an accuracy of 80% for classifying tongue/non-tongue related collapse. Our results indicate that the audio signal recorded during sleep can be helpful in identifying the site-of-collapse and therefore could potentially be used as a new tool for deciding appropriate treatment for OSA.

A preliminary study of identifying the site of upper airway collapse in osa using snoring signals

A preliminary study of identifying the site of upper airway collapse in osa using snoring signals

Identifying the site of upper airway collapse in OSA patients using snoring signal

Identifying the site of upper airway collapse in OSA patients using snoring signal

A Preliminary Study of the Automatic Classification of the Site of Airway Collapse in OSA patients Using Snoring Signals.

In this study we investigated if audio signals may carry information related to the site of obstruction of the upper airway. The information regarding the site of collapse could improve obstructive sleep apnoea (OSA) treatment by allowing more individualized or structure-specific therapy. In this preliminary study, we developed an algorithm for automatically determining the site of collapse in 13 OSA patients through snoring analysis. Audio was recorded with a ceiling mounted microphone, simultaneously with full-night polysomnography during sleep. The surrogate measure of the site of airway collapse was identified by manual analysis of the nasal pressure signal. We extracted various time and frequency features of audio signal to classify the signal into "lateral wall", "palate" and "tongue base" related collapse. The classification was carried out with a Gaussian mixture model classifier. Performance of the proposed model showed that it can achieve an overall accuracy of 78.9±0.96% with specificity and PPV of (89.3±0.81%, 78±1.5%) (73.2±1.3% , 83.2±1.8%) (61.3±2.5%, 71.8±1.3%) for lateral wall, palate and tongue base related collapse respectively. Our preliminary results suggest that the audio signal may be helpful in identifying the site of obstruction and therefore maybe a useful tool for deciding appropriate therapy.