Rees Catherine J, Blalock P David, Kemp Shannon E, Halum Stacey L, Koufman Jamie A
Department of Otolaryngology-Head and Neck Surgery UC Davis, Sacramento.
Otolaryngol Head Neck Surg. 2007 Oct;137(4):576-81. doi: 10.1016/j.otohns.2007.03.040.
To determine the utility of spectral analysis in the differentiation of adductor-type spasmodic dysphonia (AdSD) from muscle tension dysphonia (MTD).
Prospective blinded study.
Forty-seven samples of AdSD-connected speech spectrograms from 27 subjects and 17 samples of MTD-connected speech spectrograms from 15 subjects were selected from clinical charts and de-identified. These spectrograms were reviewed independently and blindly by two speech language pathologists experienced in spectrography. The speech language pathologists designated the spectrogram as consistent with AdSD and MTD, and these designations were compared with actual clinical diagnoses.
The ability to differentiate AdSD from MTD with spectral analysis was 94% for rater #1 and 98% for rater #2. No MTD subjects were incorrectly diagnosed as having SD (100% specificity).
This study suggests that experienced speech language pathologists can distinguish AdSD from MTD with a high degree of sensitivity and specificity based on spectral analysis. Spectral analysis is especially useful in cases where perceptual analysis and clinical evaluation alone are insufficient.
确定频谱分析在鉴别内收型痉挛性发音障碍（AdSD）与肌张力障碍性发音障碍（MTD）中的作用。
前瞻性盲法研究。
从临床病历中选取27名受试者的47份与AdSD相关的语音频谱图样本以及15名受试者的17份与MTD相关的语音频谱图样本，并对其进行去识别处理。两位有光谱分析经验的言语病理学家独立且盲法地审查这些频谱图。言语病理学家将频谱图判定为与AdSD和MTD一致，并将这些判定结果与实际临床诊断进行比较。
评分者1通过频谱分析鉴别AdSD与MTD的能力为94%，评分者2为98%。没有MTD受试者被错误诊断为痉挛性发音障碍（SD）（特异性为100%）。
本研究表明，经验丰富的言语病理学家基于频谱分析能够以高度的敏感性和特异性区分AdSD与MTD。在仅靠感知分析和临床评估不足的情况下，频谱分析尤其有用。
Zotero 插件