Snoring Sounds Research Articles

Relationship Between Snoring Sound Intensity and Sleepiness in Patients with Obstructive Sleep Apnea

Subjectively assessed snoring and sleepiness are known to be related. However, no evidence supporting the usefulness of snoring measurements exists. We examined whether the objectively measured snoring intensity was correlated with sleepiness. The records of 515 patients who underwent polysomnography for suspected obstructive sleep apnea were retrospectively reviewed. Subjective sleepiness was assessed using the Epworth sleepiness scale (ESS). Snoring intensity was assessed using the highest one percentile ambient sound pressure level (L1) attained while asleep during polysomnography. L1 was correlated with ESS in apneic patients with an apnea-hypopnea index (AHI) > or =15 (r = 0.38, p <0.0001), but not in other patients. The correlation in the apneic patients was preserved after adjustments for various confounding factors, including the AHI. A step-wise multiple regression in the apneic patients adopted desaturation time, L1, daily sleep time, subjective snoring, and nasal obstruction symptoms as determinants for the ESS. L1 was correlated with the mean pulse rate during polysomnography but not with sleep fragmentation variables after adjustment for the AHI. The measured snoring intensity was independently related to sleepiness in apneics. Snoring intensity may explain part of sleepiness that cannot be fully explained by ordinary polysomnographic variables.

Frequency analysis of snoring sounds during simulated and nocturnal snoring

Despite several ways of investigation, such as clinical examination, drug-induced sleep endoscopy and pressure measurement of the upper airway, it is still difficult to locate the site of snoring exactly. Frequency analysis of snoring sounds is described as a promising diagnostic tool. The aim of the study was to examine simulated snoring under conditions awake, record the produced snoring sounds and compare those sounds with nocturnal snoring. A total of 50 snoring male patients were examined clinically by flexible nasal endoscopy and simulated snoring under conditions awake, and the simulated snoring sounds were recorded. Additionally, nocturnal snoring sounds were recorded during nighttime polysomnography. Snoring events were analyzed by fast-fourier-transformation and the intensity peaks 1-5 were evaluated. Rhythmic and non-rhythmic snoring events were distinguished depending on present obstructive apneas. Clinical and polysomnographical data were correlated with the results of the frequency analysis of the snoring sounds. Simulated snoring sounds revealed a low frequency of 200 Hz in intensity peaks 1 and 2 with an increase up to 3,000 Hz in peaks 3-5. Similar frequency patterns were detected in rhythmic nocturnal snoring. Non-rhythmic snoring events revealed frequency patterns between 2,000 and 3,000 Hz in all five intensity peaks. Simulated snoring resembles rhythmic nocturnal snoring with low-frequency intensity peaks, whereas non-rhythmic snoring revealed high frequencies. The examination during simulated snoring and frequency analysis of snoring sounds might contribute in locating the pathogenesis of snoring.

Bench study of auto-CPAP devices using a collapsible upper airway model with upstream resistance

The aim of this study was to investigate the response of auto-CPAP devices to respiratory events (apnea, hypopnea, flow-limitation and snoring) on the same condition using a physiological upper airway model. The hypothesis of this study is that collapsibility of the flow-limiting collapsible segment of the airway is influenced by the upstream airway resistance. Five auto-CPAP devices, AutoSet T, AutoSet Spirit, Goodnight 420E, PV10i and REMstar Auto were evaluated. Apnea: all the devices increased the auto-CPAP level, while AutoSet T and AutoSet Spirit did not respond to apnea for 30s. Hypopnea: all the devices except the AutoSet T and Goodnight 420E increased pressure. Flow-limitation: all the devices except the PV10i and REMstar Auto increased pressure. Snoring: the snoring sounds disappeared when REMstar Auto and PV10i were used, and the Goodnight 420E lowered the level of snoring. In conclusion, the response of auto-CPAP devices to respiratory events differed. Collapsible upper airway model with upstream resistance is useful for the first-step assessment of auto-CPAP devices.

Analysed snoring sounds correlate to obstructive sleep disordered breathing

Snoring occurs as a major symptom in patients with sleep disordered breathing (SDB). The aetiology of snoring remains still unclear despite various attempts to localize snoring. The correlation between different snoring sounds and the severity of SDB has not yet been investigated in a larger population. The aim of this study was to record and analyse snoring sounds and to correlate the obtained data with clinical and polysomnographical parameters. Sixty male patients with suspected SDB and reported snoring underwent a clinical examination and night time polysomnography. The parallel digitally recorded snoring sounds were analysed by fast fourier transformation (FFT). Peak intensity was determined from the power spectrum. The periodicity of snoring was classified into rhythmic and non-rhythmic snoring according to the presence of air flow interruptions due to obstructive apneas. Patients with primary snoring revealed peak intensities between 100 and 300 Hz. Patients with an obstructive sleep apnea syndrome (OSAS) revealed peak intensities above 1,000 Hz. Polysomnographical data (AHI, mean and minimum SpO(2)) as well as body mass index (BMI) correlated with peak intensity of the power spectrum. None of the parameters of the clinical examination correlated with peak intensity. Frequency analysis of snoring sounds provides a useful diagnostic tool to distinguish between different patterns of snoring and respective SDB. The topodiagnosis of snoring is not possible by means of frequency analysis or clinical examination alone. Acoustical analysis of snoring sounds seems a promising additional diagnostic tool to verify different types of SDB in snoring patients.

Snoring noise pollution—the need for objective quantification of annoyance, regulatory guidelines and mandatory therapy for snoring

Habitual snoring without episodes of apnea or hypoventilation and without respiratory related arousals is considered to be annoying and without any need for treatment. However, studies seem to suggest an enormous psychosocial impact of annoyance for the bed partner. Apart from subjective questionnaires there still exists no generally accepted mode of measurement that can describe snoring objectively. We therefore adapted methods developed for environmental medicine and established a new snore score using psycho-acoustic parameters. For quantification of snoring noise we conducted nocturnal measurements in 19 habitual snorers. Free-field snore sounds were acquired with two low-cost non-contact microphones and transferred to a PC (sampling frequency 11 kHz). The data were recorded, analysed and stored automatically using a MATLAB script. Following the analysis of sound characteristics and levels, the score was computed from relevant parameters containing the rating level (LR), maximum level, two percentile levels for frequent maxima (L5S; L1) and snoring time. The determined values substantially exceeded the prescribed limits defined by WHO noise guidelines, and mainly affected the equivalent continuous sound exposure level, rating level and the immission standard values of brief noise peaks, whose maximum was exceeded by up to 32 dB(A). The Berlin snore score illustrated the objective acoustic annoyance on a scale from 0 to 100. It allows inter-individual comparison and objectifies the need for therapy. The clinical applicability of evaluating the reduction of snoring after surgical therapy is discussed exemplarily. The presented measuring method was found to be suitable for quantifying snoring noise and can be easily integrated into existing polysomnographic applications. In the case of habitual snoring with objective evidence of psychosocially disturbing acoustic annoyance, health fund providers should assume the costs of mandatory medical therapy.

Automatic detection, segmentation and assessment of snoring from ambient acoustic data

Snoring is a prevalent condition with a variety of negative social effects and associated health problems. Treatments, both surgical and therapeutic, have been developed, but the objective non-invasive monitoring of their success remains problematic. We present a method which allows the automatic monitoring of snoring characteristics, such as intensity and frequency, from audio data captured via a freestanding microphone. This represents a simple and portable diagnostic alternative to polysomnography. Our system is based on methods that have proved effective in the field of speech recognition. Hidden Markov models (HMMs) were employed as basic elements with which to model different types of sound by means of spectrally based features. This allows periods of snoring to be identified, while rejecting silence, breathing and other sounds. Training and test data were gathered from six subjects, and annotated appropriately. The system was tested by requiring it to automatically classify snoring sounds in new audio recordings and then comparing the result with manually obtained annotations. We found that our system was able to correctly identify snores with 82–89% accuracy, despite the small size of the training set. We could further demonstrate how this segmentation can be used to measure the snoring intensity, snoring frequency and snoring index. We conclude that a system based on hidden Markov models and spectrally based features is effective in the automatic detection and monitoring of snoring from audio data.

A Snore Extraction Method from Mixed Sound for a Mobile Snore Recorder

This paper presents a snore recorder that can separate snores from their delayed mixtures. This is useful to study the snore sounds of individuals when these sounds occur in a normal in-home sleeping environment, where two people are sleeping together and both produce sounds. Based on methods for blind source separation, we give a snore separator that solves the blind delayed source separation problem and provide a performance index to monitor its convergence. The separated snores can be analyzed to detect symptoms of sleep apnea prior to polysomnography or as a monitoring device after polysomnography has been performed. Experimental results show good performance of the snore separator.

Acoustic Analysis of Snoring Sounds by a Multidimensional Voice Program

This prospective study aimed to determine whether the acoustic characteristics of snoring sounds differed between simple snorers and patients with obstructive sleep apnea syndrome (OSAS) by using a multidimensional voice program (MDVP) that analyzes various aspects of voice. Fifty-eight patients (48 men, 10 women) with a history of snoring were included in the study. All patients underwent conventional polysomnography (PSG). Twelve subjects were diagnosed as simple snorers and 46 subjects were diagnosed with OSAS. The mean body mass index (BMI) of simple snorers was 24.7 kg/m and that of patients with OSAS was 25.8 kg/m. Natural overnight snoring was recorded from each subject while they slept during PSG. Using the multiple token protocols of MDVP, 30 snores from each subject were analyzed automatically. For data analysis, four markers were used: peak frequency, soft phonation index (SPI), noise to harmonics ratio (NHR), and power ratio. The Mann-Whitney U test revealed significant differences between the SPI, NHR, and power ratio of simple snorers and patients with OSAS. Simple snorers had a high SPI value. OSAS-related snorers demonstrated a high NHR and low power ratio. MDVP can be used for snoring sound analysis as a noninvasive examination of sleep-related breathing disorders for differential diagnosis. However, a suitable option that is rapid and has an easy-to-use interface would be more advantageous for analyzing snoring sounds.

Pitch jump probability measures for the analysis of snoring sounds in apnea

Obstructive sleep apnea (OSA) is a highly prevalent disease in which upper airways are collapsed during sleep, leading to serious consequences. The gold standard of diagnosis, called polysomnography (PSG), requires a full-night hospital stay connected to over ten channels of measurements requiring physical contact with sensors. PSG is inconvenient, expensive and unsuited for community screening. Snoring is the earliest symptom of OSA, but its potential in clinical diagnosis is not fully recognized yet. Diagnostic systems intent on using snore-related sounds (SRS) face the tough problem of how to define a snore. In this paper, we present a working definition of a snore, and propose algorithms to segment SRS into classes of pure breathing, silence and voiced/unvoiced snores. We propose a novel feature termed the ‘intra-snore-pitch-jump’ (ISPJ) to diagnose OSA. Working on clinical data, we show that ISPJ delivers OSA detection sensitivities of 86–100% while holding specificity at 50–80%. These numbers indicate that snore sounds and the ISPJ have the potential to be good candidates for a take-home device for OSA screening. Snore sounds have the significant advantage in that they can be conveniently acquired with low-cost non-contact equipment. The segmentation results presented in this paper have been derived using data from eight patients as the training set and another eight patients as the testing set. ISPJ-based OSA detection results have been derived using training data from 16 subjects and testing data from 29 subjects.

Acoustic analysis of snoring sounds with chaos theory

Snoring sounds are approximately periodic waves with noise. Therefore, the characteristics of snoring sound cannot be represented appropriately by frequency analysis. The snoring sounds of six patients with sleep apnea syndrome were evaluated using the chaos theory. To evaluate the chaotic nature of snoring sound, phase space, Poincare section, and time evolutions of Lyapunov exponents were calculated. At the same time, the snoring sound was analyzed by the conventional FFT method. Analytical results were as follows. An attractor of snoring sound was drawn in a four-dimensional space using the embedding theorem of Takens. The values of Lyapunov exponent (one of the indices of chaotic nature) were positive, suggesting that snoring sound demonstrates the characteristics of chaos. These nonlinear dynamic methods are expected to provide useful information for better understanding of irregular snoring sounds.

Acoustic analysis of pediatric snoring

Videotaping was conducted at home, in cooperation with patients' mothers. The recorded sounds were then analyzed. Compared with snoring sounds obtained via polysomnography, representative cases that demonstrate the usefulness of videotaping at home were presented. The subjects were eight children who received polysomnography, recording the snoring at our department during the 1-year period, from March 2002 to February 2003. The snoring recording was conducted using polysomnography. As for the taping method, the sounds from the microphone were taken into an acoustic analysis system II (TDT) and recorded on an MD, using a ZS-M75 (SONY). As for the method of analysis, the MD-recorded snore sounds were analyzed using Time Frequency Representation (TFR) (Avaaz Innovation), an acoustic analysis software program for Windows. The snoring sounds recorded by home video camera as well as those obtained by polysomnography were analyzed. The snoring sounds obtained via polysomnography and by the home video camera contained noise components on sound spectrograms and showed waves in which the basic frequency changed. It thought that the noise made by the adenoids and the hypertrophy of palatine tonsil were not the cause of snoring, but rather that the vibration of the soft palate and the uvula participated in pediatric snoring.

Sound frequency analysis and the site of snoring in natural and induced sleep.

The aim of this study was to compare the snoring sounds induced during sleep nasendoscopy, and to compare them with those of natural sleep using sound frequency spectra. The snoring of 16 subjects was digitally recorded during natural and induced sleep, noting the site of vibration during sleep nasendoscopy. Patients with palatal snoring during sleep nasendoscopy had a median peak frequency at 137 Hz (118 snore samples). The peak frequency of tongue-base snoring was 1243 Hz (10 snore samples), and simultaneous palate and tongue was 190 Hz (six snore samples). The median power ratios were 7, 0.2 and 5 respectively. The centre frequencies were 371, 1094 and 404 Hz respectively. Epiglottic snores had a peak frequency of 490 Hz (five snore samples). Comparison of the induced (n = 118) and natural (n = 300) snore samples of the 12 palatal snorers showed a significant difference in both the power ratio and centre frequencies (P = 0.031 and P = 0.049). The peak frequency position was similar (P = 0.34). Our results indicate that induced snores contain a higher frequency component of sound, not evident during natural snoring. This is consistent with an element of tongue-base snoring. Although there is good correlation generally, sleep nasendoscopy may not accurately reflect natural snoring.

Palatal snoring identified by acoustic crest factoranalysis

The differentiation of palatal from non-palatal snoring is very important for ENT surgeons trying to determine whether palatal surgery would be curative. At present sleep nasendoscopy is the accepted method. Palatal vibration produces marked modulation of sound loudness at low frequency (below ~100 Hz). We calculate a crest factor for the sound waveform (ratio of peak to root mean square (rms) value in any given epoch), as a measure of the degree of modulation. Free-field snore sounds were recorded from 11 supine adult patients under intravenous sedation (midazolam), using a digital tape recorder. Recordings were transferred to a PC (sampling frequency 11 kHz), and analysed using code written by us. Direct visual confirmation of the site of snoring was gained from simultaneous sleep nasendoscopy, taken as the gold standard. In six patients the dominant site was the soft palate. The non-palatal group (five patients) comprised one epiglottic, two hypopharyngeal and two tongue base snorers. The crest factor was found to be significantly higher for palatal snorers (p<0.01, Student-t or Mann-Whitney tests). Furthermore, palatal could be distinguished from non-palatal snorers on the basis of crest factor alone in all 11 cases, making this a promising non-invasive diagnostic technique.

An acoustic screening test for obstructive sleep apnoea

Third-octave sound analysis was performed on the snoring sounds of nine subjects with obstructive sleep apnoea (OSA) and 18 with simple snoring. Both groups demonstrated a large low frequency peak in linear sound levels at around 80 Hz. However, the OSA group displayed a substantially larger high frequency sound component. We utilized this fact in the development of an acoustic index (Hawke Index: HI) which describes the ratio between the overall A-weighted and linear sound levels for the recorded snoring sound of each subject [HI = dB(A)/dB(SPL) for Lmax]. There was a significant positive correlation between the apnoea/hypopnoea index and the HI (r = 0.73, t = 5.3, 25df, P < 0.001). If a value of 0.90 or greater was taken as diagnostic of OSA, the HI exhibited 67% sensitivity, 100% specificity, 100% positive and 86% negative predictive accuracy. With further development, we believe this acoustic phenomena may have a role as a screening test in the diagnosis of OSA.

Digital Monitoring of Sleep-disordered Breathing Using Snoring Sound and Arterial Oxygen Saturation

A new portable digital recorder (SNORESAT) that uses the sound of snoring and arterial oxygen saturation (SaO2) to monitor breathing abnormalities during sleep was constructed and compared in the laboratory with standard overnight polysomnography (PSG). The device digitally records sound from a transducer applied to the chest and SaO2 from a commercially available ear oximeter. A snore is identified when the moving time average of the sound exceeds a threshold voltage level longer than 0.26 s. The stored data are transferred to a personal computer for poststudy analysis. An analysis algorithm identifies a respiratory disturbance event when a quiet period of 10 to 120 s separates two snores and is associated with a fall in SaO2 exceeding 3%. The respiratory disturbance index (RDI), mean apnea duration, mean lowest SaO2, and number of desaturations > 3% are computed. A total of 129 referrals to the sleep apnea outpatient clinic underwent simultaneous all-night recording of PSG and SNORESAT. Using the computed RDI recorded by the SNORESAT, the sensitivity and specificity of the monitor in detecting sleep apnea syndrome (SAS) ranged between 84 and 90% and 95 and 98%, respectively, depending on the PSG value of RDI used to define SAS (range, > or = 7 to > or = 20 events/h). Using a PSG value of RDI > or = 10, or > or = 20 RD/h as the definition for SAS, the prevalence of SAS in the referral population was 45 and 31%, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

いびきの音響分析 基本周波数について

We attempted to identify the sound source of snoring only by analyzing snoring sounds. Snoring sounds were tape-recorded at home and brought to the hospital by the patients, and with a sound spectrograph and a signal processor we analyzed the results and compared them with those of patients whose sources of snoring were confirmed fiberscopically.1. Patients whose fundamental frequency (FF) was fairly regular, the FFs ranged from 15 Hz to 300Hz, usually under 170Hz.2. In some cases, the FF changed in a single snore. A decrease of FF was observed more often than a increase.3. The pitch perturbation quotients of ordinary snoring sounds were larger than those of vocal fold vibrations.4. Among the sound sources confirmed fiberscopically, the great majority were vibrations of the soft palate, the FFs ranging from 20 Hz to 200 Hz. The few other sound sources of snoring comfirmed were: vocal folds (210 to 300Hz), the arytenoids (130 to 340Hz), the epiglottis (45 to 90Hz), and the tongue base (40 to 50HZ).

Digital signal analysis of snoring sounds in children

Snoring sounds of 5 children with sleep-dependent compromised breathing were recorded. Underlying diseases were large adenoids, adeno-tonsillar hyperplasia, hemangioma of the tongue-base and Rubinstein-Taybi's syndrome associated with micrognathia. For comparison snoring sounds of a 42-year-old chronic snorer were analyzed. Time series of the power spectral density of the breath sounds were calculated as well as the mean power spectra of the recordings. The results show that the frequency spectrum reflects the pathomechanism responsible for the production of the abnormal breath sound. In adults, simple snoring is due in large part to vibrations of the soft palate. This can be identified in the frequency spectrum as a low-frequency component with a large number of harmonics. In children with enlarged adenoids and tonsils the soft palate is impeded in its movement which can be demonstrated in the frequency spectrum as a lack of low-frequency components and harmonics. Craniofacial anomalies are characterized by special spectral patterns.

Device for preventing of snoring

A device for preventing snoring of a sleeping person comprises an arrangement (5-9) for detecting snoring sounds and an apparatus (17) controlled by this arrangement and adapted to influence the person to stop snoring. The arrangement has a sound-receiving microphone (5) and at least one frequency filter (6, 8), arranged to deliver signals deriving from sounds with frequencies typical for snoring sounds in order to determine if the present sound derives from snores. The arrangement further comprises a circuit (10-14) for determining if the signals delivered by the microphone are periodically appearing at time intervals, which are typical for snores. The circuit comprises a counter (13) which is arranged to count the number of snores. The arrangement (5-14) of the device is adapted to send activating control pulses to the apparatus when the counter (13) has detected a predetermined number of successive snores with the time intervals.

How can one recognize a velum snorer?

In ten habitually snoring subjects and ten patients with obstructive sleep apnea syndrome, preoperative snoring sounds above a preset sound pressure level were recorded and a frequency spectrum analysis by means of time series Fast Fourier Transformation was performed. All patients underwent uvulopalato-pharyngoplasty after which the snoring sounds were recorded again and analysed under the same conditions. Preoperative frequency spectra in heavy snorers were dominated by a component which results from the movement of the soft palate and uvula and which presents as a low-frequency vibration of 25 to 50 c/s and its harmonics up to several hundred c/s. After UPPP the sound pressure level of the snoring noise in heavy snorers was reduced by 20 to 30 decibels and could not be recorded any more. The frequency spectrum of snoring noises in patients with obstructive sleep apnea syndrome differed from those of heavy snorers. In apnea patients the snoring noise usually consisted of other and higher frequency components which masked the vibration of the soft palate and uvula to a more or less high degree. An extreme degree of masking is reached in snoring noises following apneic episodes when identification of the movement of the soft palate is impossible. After performing UPPP in an apneic patient the low-frequency content of the spectrum below about 400 Hz was reduced dramatically but higher frequency parts were still present. This resulted in only minor reduction of snoring noise sound pressure level. Our results are in agreement with other authors in showing that UPPP is not successful when the site of obstruction is not in the velo-pharyngeal plane.(ABSTRACT TRUNCATED AT 250 WORDS)

Cognitive Function in Middle-aged Snorers and Controls: Role of Excessive Daytime Somnolence and Sleep-Related Hypoxic Events

Association of snoring and cognitive function was studied in 46 habitually snoring men ages 41-52 years, and 60 occasionally or never-snoring control male subjects of the same age group. Sleep recordings with monitoring of apneas and hypopneas were made with the static-charge sensitive bed method. Blood oxygen saturation was measured with an oximeter and the snoring sounds were recorded with a microphone after clinical and neuropsychological assessment. A questionnaire with items on excessive daytime somnolence (EDS), sleep, and snoring quality was also used. EDS (as measured by items on the questionnaire) associated with tests requiring concentration, memory retention, and verbal and spatial skills in the habitual snorers group. The number of oxygen desaturation episodes exceeding 4% associated with defective delayed Recall of Logical Stories of the Wechsler Memory Scale and with spatial orientation (Clock test) in the habitual snorers' group even after adjusting for age and obesity.