The upper airway (UA) is in general thicker and narrower in obstructive sleep apnea (OSA) population than in normal. Additionally, the UA changes during sleep are much more in the OSA population. The UA changes can alter the tracheal breathing sound (TBS) characteristics. Therefore, we hypothesize the TBS changes from wakefulness to sleep are significantly correlated to the OSA severity; thus, they may represent the physiological characteristics of the UA. To investigate our hypothesis, we recorded TBS of 18 mild-OSA (AHI < 15) and 22 moderate/severe-OSA (AHI > 15) during daytime (wakefulness) and then during sleep. The power spectral density (PSD) of the TBS was calculated and compared within the two OSA groups and between wakefulness and sleep. The average PSD of the mild-OSA group in the low-frequency range (< 280 Hz) was found to be decreased significantly from wakefulness to sleep (p-value < 10⁻⁴). On the other hand, the average PSD of the moderate/severe-OSA group in the high-frequency range (> 900 Hz) increased marginally significantly from wakefulness to sleep (p-value < 9 × 10⁻³). Our findings show that the changes in spectral characteristics of TBS from wakefulness to sleep correlate with the severity of OSA and can represent physiological variations of UA. Therefore, TBS analysis has the potentials to assist with diagnosis and clinical management decisions in OSA patients based on their OSA severity stratification; thus, obviating the need for more expensive and time-consuming sleep studies.

Tracheal breathing sound (TBS) changes from wakefulness to sleep and their correlation with Obstructive sleep apnea (OSA) were investigated in individuals with different levels of OSA severity. We also assessed the classification power of the spectral characteristics of these TBS for screening purposes. Consequently, we analyzed and compared spectral characteristics of TBS recorded during wakefulness (a combination of mouth and nasal TBS) to those during sleep for mild and moderate/severe OSA groups.

阻塞性睡眠呼吸暂停（OSA）人群的上呼吸道（UA）通常比正常人更粗，更窄。另外，睡眠期间UA的变化在OSA人群中更为明显。UA的改变会改变气管呼吸音（TBS）的特性。因此，我们假设从清醒到睡眠的TBS变化与OSA严重程度显着相关。因此，它们可以代表UA的生理特征。为了研究我们的假设，我们在白天（清醒）然后睡眠期间记录了18种轻度OSA（AHI <15）和22种中度/重度OSA（AHI> 15）的TBS。计算了TBS的功率谱密度（PSD），并在两个OSA组中以及清醒和睡眠之间进行了比较。轻度OSA组在低频范围内的平均PSD（<p-值＜10 ^-4）。另一方面，中度/重度OSA组在高频范围（> 900 Hz）中的平均PSD从清醒到睡眠都显着增加（p值<9×10 ^-3）。我们的发现表明，从清醒到睡眠，TBS光谱特征的变化与OSA的严重程度相关，并且可以代表UA的生理变化。因此，TBS分析有可能根据OSA严重程度分层来协助OSA患者的诊断和临床管理决策。因此，无需进行更昂贵和耗时的睡眠研究。

在具有不同OSA严重性水平的个体中，研究了气管呼吸声（TBS）从清醒到睡眠的变化及其与阻塞性睡眠呼吸暂停（OSA）的相关性。我们还评估了这些TBS光谱特征的分类能力，以进行筛选。因此，对于轻度和中度/重度OSA组，我们分析并比较了清醒（嘴巴和鼻腔TBS组合）与睡眠期间所记录的TBS光谱特征。