Hosseinifar Shamim, Torabinezhad Farhad, Ghelichi Leila, Roudbari Masoud, Silverman Erin Pearson, Faham Maryam
Department of Speech and Language Pathology, School of Rehabilitation Sciences, Iran University of Medical Sciences, Tehran, Iran.
Antimicrobial Resistance Research Center, Rasoul-e-Akram Hospital, Department of Biostatistics, School of Public Health, Iran University of Medical Sciences, Tehran, Iran.
J Voice. 2018 Nov;32(6):705-709. doi: 10.1016/j.jvoice.2017.08.015. Epub 2017 Oct 21.
Perceptual and acoustic analyses are essential tools that help voice therapists comprehensively assess voice quality. While perceptual evaluations are subjective and are influenced by external and culturally driven factors, acoustic analysis is an objective and reliable means of evaluating voice. The goals of this study were (1) to determine which acoustic parameters were predicted by perceptual voice quality and (2) to assess the effect of a short period of training on the reliability of perceptual voice analyses for Persian speakers.
This was a cross-sectional study. Subjects were 20 patients with various voice disorders. Voice samples were obtained during text reading and /a/ prolongation. Fifteen expert voice clinicians completed perceptual evaluations on voice samples using the Grade, Roughness, Breathiness, Asthenia, and Strain scale. We repeated this process after a short period of perceptual voice evaluation training. Acoustic analysis was completed using the Praat program. We used the intraclass correlation coefficient (ICC) for reliability measurement of the perceptual evaluation results and ordinal regression procedures to analyze all data. Significance level was set at P < 0.05.
Both intrarater and interrater reliability increased after training, for all five parameters. The ICC for grade increased to 0.95 after training. Grade and roughness significantly predicted fundamental frequency (F0) (P = 0.021 and P = 0.030, respectively) and harmonic-to-noise ratio (HNR) (P = 0.019 and P = 0.016, respectively). Breathiness significantly predicted shimmer (P = 0.013).
Training had a positive effect and increased the reliability of perceptual voice evaluation. For Persian listeners, changes in F0, increases in HNR, and shimmer were perceptually associated with poor voice quality.
感知分析和声学分析是帮助嗓音治疗师全面评估嗓音质量的重要工具。虽然感知评估具有主观性，会受到外部和文化驱动因素的影响，但声学分析是评估嗓音的一种客观且可靠的方法。本研究的目的是：（1）确定哪些声学参数可由感知嗓音质量预测；（2）评估短期训练对波斯语使用者感知嗓音分析可靠性的影响。
这是一项横断面研究。研究对象为20名患有各种嗓音障碍的患者。在文本朗读和/a/延长过程中采集嗓音样本。15名专业嗓音临床医生使用等级、粗糙度、气息声、无力感和紧张度量表对嗓音样本进行感知评估。在进行短期的感知嗓音评估训练后，我们重复了这一过程。使用Praat程序完成声学分析。我们使用组内相关系数（ICC）来衡量感知评估结果的可靠性，并使用有序回归程序分析所有数据。显著性水平设定为P < 0.05。
经过训练后，所有五个参数的评分者内信度和评分者间信度均有所提高。训练后等级的ICC增至0.95。等级和粗糙度显著预测了基频（F0）（分别为P = 0.021和P = 0.030）和谐波噪声比（HNR）（分别为P = 0.019和P = 0.016）。气息声显著预测了微扰（P = 0.013）。
训练产生了积极影响，提高了感知嗓音评估的可靠性。对于波斯语听众而言，F0的变化、HNR的增加以及微扰在感知上与嗓音质量差相关。
