Could formant frequencies of snore signals be an alternative means for the diagnosis of obstructive sleep apnea?

Investigation of Obstructive Sleep Apnea Using Nonlinear Mode Interactions in Nonstationary Snore Signals

“…AHI refers to apnea-hypopnea index in events/h; AUC, area under receiver operating characteristic curve; p, value of statistical significance which was considered to be present when p < 0.05; Sens, sensitivity; Spec, specificity; M, male; F, female; C, both males and females combined; A, apneic; B, benign. 38,53 In contrast to the quadratic model, the exponential and the power models perform better, in terms of RSD values (0.4347-0.5277 events/h) and AHI prediction (8-10.7 events/h), despite having lower R 2 (0.2911-0.5189). Since R 2 is not the major criterion for judging whether a fit is practical, 34 one may infer that the latter two regression models could realistically interpret the relationships between the severity of OSA and the potential dominant frequency modes involved in the nonlinear interactions in snore signals.…”

“…These techniques include, but are not limited to, analyzing heart rate variability, 24 thoracic movement, 43 and acoustical properties of snores. 1,2,8,11,17,19,22,30,37,38,40,46,47,53 Among these techniques, snore-based analysis has received growing interests as it can possibly offer safe and non-invasive data measurement with minimum labor and medical costs.…”

“…45,57 Motivated by the impact of structural changes in the UA on the acoustical properties of snores, various methodologies for identifying patients with or without OSA using snore signals have been investigated. These methodologies involve counting number of snores, 22 estimating sound intensity levels, 46,53 computing power spectral densities, 8,17,19 recognizing pitch accents, 2,47 modeling formant frequencies, 11,30,38 extracting phase couplings through Fourier-based bispectral approach, 1,30,40 and even measuring psychoacoustic metrics. 37 While lending support to the hypothesis that snore signals could likely be an alternative means for assessing OSA, these methodologies are mostly tailored to the framework of classical linear models and ergodic Gaussian random processes.…”

Role of Upper Airway Dimensions in Snore Production: Acoustical and Perceptual Findings

“…16 In the light of snoring (a hallmark of OSA), fluttering vibrations of soft tissues and/or noise-like turbulent airflow at constrictions inevitably produce acoustic waves, which are then spectrally modified by the UA anatomical structures (e.g., cross-sectional airway dimensions) to create distinct sounds before reaching the listener. 10,30 In recent years, multiple acoustic markers of snore signals have been proposed to discriminate between apneic and benign patients, with a common interest to develop a non-invasive, inexpensive, and rapid screening tool for OSA. These markers include, but are not limited to, first formant frequency (F1) in the linear prediction (LP) spectrum 30 ; peak frequency (PF) in the fast Fourier transformation curve 25 ; frequency modes in the bispectrum 32 ; mean and standard deviation of the coefficient of variation in a short signal frame 7 ; soft phonation index and noise-to-harmonics ratio 17 ; and even psychoacoustic metrics in terms of loudness, sharpness, roughness, fluctuation strength, and annoyance.…”

“…10,30 In recent years, multiple acoustic markers of snore signals have been proposed to discriminate between apneic and benign patients, with a common interest to develop a non-invasive, inexpensive, and rapid screening tool for OSA. These markers include, but are not limited to, first formant frequency (F1) in the linear prediction (LP) spectrum 30 ; peak frequency (PF) in the fast Fourier transformation curve 25 ; frequency modes in the bispectrum 32 ; mean and standard deviation of the coefficient of variation in a short signal frame 7 ; soft phonation index and noise-to-harmonics ratio 17 ; and even psychoacoustic metrics in terms of loudness, sharpness, roughness, fluctuation strength, and annoyance. 29 Although the snore-driven markers appear to shed light on OSA detection, there is little research on correlation between the UA dimensions and the properties of snores.…”