论文部分内容阅读
目的探讨Lyapunov系数在嗓音障碍客观分析中的意义,完善嗓音障碍的客观测试方法。方法声音样本采自132例嗓音障碍患者和47例嗓音正常者。采用法国开发的“EVA”嗓音工作站,选择2秒平稳段长元音/α/声信号作为分析样本,测试基频微扰(jitter)和计算Lyapunov系数。主观听感知评估采用日本言语音声学会制定的嗓音障碍评估系统GRBAS中的总嘶哑度(grade,G),4级评估标准。结果Lyapunov系数在不同的嗓音障碍分级组间均有统计学差异,jitter只在中度和重度嗓音障碍组间有统计学差异;Lyapunov系数与听感知评估结果的一致性达到46%,jitter的一致性为36%。回归分析结果显示,Lyapunov系数和jitter之间既有相关性,但又有不同(r2=0.582,P<0.001)。结论Lyapunov系数对嗓音障碍的测试优于jitter;非线性动态系统(相位图)的引入,为嗓音的客观分析开辟了新的研究领域。
Objective To explore the significance of Lyapunov coefficient in the objective analysis of voice disorders and to improve the objective test method of voice disorders. Methods Sound samples were collected from 132 patients with voice disorders and 47 patients with normal voice. Using the “EVA” voice workstation developed in France, a 2-sec stationary long vowel / α / acoustic signal was selected as an analysis sample to test the fundamental jitter and calculate the Lyapunov coefficients. The subjective listening perception was assessed using the grade (G) and grade 4 assessment criteria of GRBAS, a voice disorder assessment system developed by the Japanese Speech Acoustic Society. Results The Lyapunov coefficients were significantly different among the different voice impairment classification groups. The jitter only had significant difference between the moderate and severe voice disorders group. The consistency between the Lyapunov coefficient and the auditory perception assessment was 46% and jitter was consistent Sex is 36%. Regression analysis showed both Lyapunov coefficient and jitter were correlated, but different (r2 = 0.582, P <0.001). Conclusion The Lyapunov coefficient is superior to jitter for the detection of voice disturbances. The introduction of nonlinear dynamic system (phase diagram) opens up new research fields for the objective analysis of voice.