Yadollahi Azadeh, Rudzicz Frank, Montazeri Aman, Bradley T Douglas
Annu Int Conf IEEE Eng Med Biol Soc. 2013;2013:2924-7. doi: 10.1109/EMBC.2013.6610152.
Obstructive sleep apnea (OSA) is a common disorder due to recurrent collapse of the upper airway (UA) during sleep that increases the risk for several cardiovascular diseases. Recently, we showed that nocturnal fluid accumulation in the neck can narrow the UA and predispose to OSA. Our goal is to develop non-invasive methods to study the pathogenesis of OSA and the factors that increase the risks of developing it. Respiratory sound analysis is a simple and non-invasive way to study variations in the properties of the UA. In this study we examine whether such analysis can be used to estimate the amount of neck fluid volume and whether fluid accumulation in the neck alters the properties of these sounds. Our acoustic features include estimates of formants, pitch, energy, duration, zero crossing rate, average power, Mel frequency power, Mel cepstral coefficients, skewness, and kurtosis across segments of sleep. Our results show that while all acoustic features vary significantly among subjects, only the variations in respiratory sound energy, power, duration, pitch, and formants varied significantly over time. Decreases in energy and power over time accompany increases in neck fluid volume which may indicate narrowing of UA and consequently an increased risk of OSA. Finally, simple discriminant analysis was used to estimate broad classes of neck fluid volume from acoustic features with an accuracy of 75%. These results suggest that acoustic analysis of respiratory sounds might be used to assess the role of fluid accumulation in the neck on the pathogenesis of OSA.
阻塞性睡眠呼吸暂停(OSA)是一种常见疾病,由于睡眠期间上气道(UA)反复塌陷,增加了多种心血管疾病的风险。最近,我们发现夜间颈部液体蓄积会使上气道变窄并易患OSA。我们的目标是开发非侵入性方法来研究OSA的发病机制以及增加其发病风险的因素。呼吸音分析是一种研究上气道特性变化的简单且非侵入性的方法。在本研究中,我们检验这种分析是否可用于估计颈部液体量,以及颈部液体蓄积是否会改变这些声音的特性。我们的声学特征包括对睡眠各阶段共振峰、音高、能量、持续时间、过零率、平均功率、梅尔频率功率、梅尔倒谱系数、偏度和峰度的估计。我们的结果表明,虽然所有声学特征在受试者之间有显著差异,但只有呼吸音能量、功率、持续时间、音高和共振峰的变化随时间有显著差异。随着时间推移能量和功率的降低伴随着颈部液体量的增加,这可能表明上气道变窄,从而OSA风险增加。最后,使用简单判别分析从声学特征估计颈部液体量的大致类别,准确率为75%。这些结果表明,呼吸音的声学分析可能用于评估颈部液体蓄积在OSA发病机制中的作用。