Cannito Michael P, Buder Eugene H, Chorna Lesya B
The University of Memphis, TN 38105, USA.
J Voice. 2005 Sep;19(3):391-410. doi: 10.1016/j.jvoice.2004.07.001.
Spectral amplitude measures are sensitive to varying degrees of vocal fold adduction in normal speakers. This study examined the applicability of harmonic amplitude differences to adductor spasmodic dysphonia (ADSD) in comparison with normal controls. Amplitudes of the first and second harmonics (H1, H2) and of harmonics affiliated with the first, second, and third formants (A1, A2, A3) were obtained from spectra of vowels and /i/ excerpted from connected speech. Results indicated that these measures could be made reliably in ADSD. With the exception of H1()-H2(), harmonic amplitude differences (H1()-A1, H1()-A2, and H1()-A3()) exhibited significant negative linear relationships (P < 0.05) with clinical judgments of overall severity. The four harmonic amplitude differences significantly differentiated between pre-BT and post-BT productions (P < 0.05). After treatment, measurements from detected significant differences between ADSD and normal controls (P < 0.05), but measurements from /i/ did not. LTAS analysis of ADSD patients' speech samples proved a good fit with harmonic amplitude difference measures. Harmonic amplitude differences also significantly correlated with perceptual judgments of breathiness and roughness (P < 0.05). These findings demonstrate high clinical applicability for harmonic amplitude differences for characterizing phonation in the speech of persons with ADSD, as well as normal speakers, and they suggest promise for future application to other voice pathologies.
频谱幅度测量对正常说话者不同程度的声带内收敏感。本研究将谐波幅度差异与正常对照组进行比较,检验其在内收型痉挛性发声障碍(ADSD)中的适用性。从连贯言语中提取的元音和/i/的频谱中获取第一和第二谐波(H1、H2)以及与第一、第二和第三共振峰相关的谐波(A1、A2、A3)的幅度。结果表明,这些测量在ADSD中可以可靠地进行。除H1()-H2()外,谐波幅度差异(H1()-A1、H1()-A2和H1()-A3())与总体严重程度的临床判断呈显著负线性关系(P<0.05)。这四个谐波幅度差异在BT前和BT后的发声之间有显著差异(P<0.05)。治疗后,从检测到的ADSD与正常对照组之间存在显著差异(P<0.05),但从/i/的测量结果没有。对ADSD患者语音样本的长时平均谱分析与谐波幅度差异测量结果拟合良好。谐波幅度差异也与呼吸音和粗糙感的感知判断显著相关(P<0.05)。这些发现表明,谐波幅度差异在表征ADSD患者以及正常说话者的语音发声方面具有很高的临床适用性,并为未来应用于其他嗓音疾病提供了前景。