Pereira Jotz Geraldo, Cervantes Onivaldo, Abrahão Márcio, Parente Settanni Flávio Aurélio, Carrara de Angelis Elisabete
Otolaryngology-Head and Neck Surgery Department at Universidade Luterana do Brasil (ULBRA), Porto Alegre, RS.
J Voice. 2002 Mar;16(1):28-31. doi: 10.1016/s0892-1997(02)00068-1.
This prospective study assessed the efficacy of computerized noise-to-harmonics ratio (NHR) to quantify perceptual and endoscopic findings of dysphonia and/or structural lesion of the vocal fold. Fifty Brazilian boys without vocal complaints were submitted to computerized, perceptual, and endoscopic examination. Thirty boys were dysphonic--3 were classified into the grade category, 5 into breathiness, 9 into roughness, and 15 into grade/breathiness. Vocal fold lesions were observed in 25 boys (17 nodules and 8 cysts). The Mann-Whitney U test revealed that NHR was significantly higher in boys with a structural lesion (p = 0.007) and in boys with dysphonia (p < 0.0001). However, according to a logistic regression model, only the occurrence of dysphonia was explained by NHR; the risk for having dysphonia increased approximately twice (odds ratio = 1.92, 95% confidence interval = 1.3-2.9) with each increase of 0.01 in NHR. Our results suggest that noise is a useful quantitative index to confirm a perceptual diagnosis of dysphonia and to evaluate quantitative changes in a dysphonic voice over time. However, we believe that computerized analysis should be used as a complement, rather than a substitute, for perceptual evaluation. Further studies with a larger sample are required to investigate the relationship between noise and lesions of the vocal folds.
这项前瞻性研究评估了计算机化噪声与谐波比率（NHR）在量化嗓音障碍的感知和内镜检查结果以及声带结构病变方面的有效性。五十名无嗓音问题的巴西男孩接受了计算机化、感知和内镜检查。三十名男孩存在嗓音障碍——3名被归类为分级类别，5名有呼吸声，9名有粗糙声，15名有分级/呼吸声。在25名男孩中观察到声带病变（17个小结和8个囊肿）。曼-惠特尼U检验显示，有结构病变的男孩（p = 0.007）和有嗓音障碍的男孩（p < 0.0001）的NHR显著更高。然而，根据逻辑回归模型，只有嗓音障碍的发生可以由NHR解释；NHR每增加0.01，出现嗓音障碍的风险大约增加两倍（优势比 = 1.92，95%置信区间 = 1.3 - 2.9）。我们的结果表明，噪声是确认嗓音障碍感知诊断以及评估嗓音障碍患者嗓音随时间的定量变化的有用定量指标。然而，我们认为计算机化分析应作为感知评估的补充，而非替代。需要进一步进行更大样本量的研究来探究噪声与声带病变之间的关系。
