Rohrmeier Christian, Herzog Michael, Ettl Tobias, Kuehnel Thomas S
Department of Otorhinolaryngology, University of Regensburg, 93042, Regensburg, Germany,
Sleep Breath. 2014 Mar;18(1):169-76. doi: 10.1007/s11325-013-0866-8. Epub 2013 Jun 21.
Although snoring is a common problem, no unequivocal definition yet exists for this acoustic phenomenon. The primary study objective was to investigate whether snoring sounds can be distinguished at all clearly from breath sounds. Our secondary objective was to evaluate whether the sound pressure level in common use and psychoacoustic parameters are suitable for making this distinction.
Twenty-five subjects exposed to 55 sound sequences were asked to decide whether these were breath sounds or snoring sounds, and to indicate how certain they were about their decision. The sound pressure level and the psychoacoustic parameters of loudness, sharpness, roughness, and fluctuation strength were then analyzed, and psychoacoustic annoyance was calculated from these parameters.
Sixteen percent of the sound sequences could not be classified unequivocally, although the individual raters stated that they were still moderately certain about their decision. The sound pressure level and psychoacoustic parameters were capable of distinguishing between breath sounds and snoring sounds. The optimum for sensitivity and specificity was 76.9 and 78.8 %, respectively.
Because snoring appears to be a subjective impression, at least in part, a generally valid acoustic definition therefore seems to be impossible. The sound pressure level and psychoacoustic parameters are suitable for distinguishing between breath sounds and snoring sounds. Nevertheless, when interpreting results, the only moderate validity of these parameters due to the absence of a universally valid definition of snoring should be taken into account.
尽管打鼾是一个常见问题，但对于这种声学现象尚无明确的定义。主要研究目的是调查打鼾声是否能与呼吸声清晰区分。次要目的是评估常用的声压级和心理声学参数是否适合进行这种区分。
让25名受试者听55个声音序列，要求他们判断这些是呼吸声还是打鼾声，并指出他们对自己判断的确定程度。然后分析声压级以及响度、尖锐度、粗糙度和波动强度等心理声学参数，并根据这些参数计算心理声学烦恼度。
16%的声音序列无法明确分类，尽管个体评分者表示他们对自己的判断仍有一定把握。声压级和心理声学参数能够区分呼吸声和打鼾声。灵敏度和特异度的最佳值分别为76.9%和78.8%。
由于打鼾似乎至少部分是一种主观印象，因此似乎不可能有一个普遍适用的声学定义。声压级和心理声学参数适合区分呼吸声和打鼾声。然而，在解释结果时，应考虑到由于缺乏打鼾的普遍有效定义，这些参数的有效性仅为中等。
