Department of Electrical and Computer Engineering, and National Centre for Audiology, University of Western Ontario, London, N6A 4B8 Ontario, Canada.
J Acoust Soc Am. 2010 Feb;127(2):1032-41. doi: 10.1121/1.3270396.
Total laryngectomy is often the treatment of choice for many individuals diagnosed with advanced laryngeal cancer. This procedure alters the normal voice production mechanism, and tracheoesophageal (TE) speech is one alternative method of voicing postlaryngectomy. TE speech is created when pulmonary air is passed through the upper esophagus to create a vibratory source that is then articulated into speech. TE speech is often characterized by abnormal voice quality. Acoustic analysis of TE speech has the potential of quantifying the voice quality and assisting the speech language pathologist in facilitating rehabilitation. Motivated in part by the recent advances in telecommunication industry for speech quality estimation, this paper investigated the application of an auditory model in predicting the ratings of TE speech by normal hearing listeners. The Moore-Glasberg auditory model was employed to extract perceptually relevant features from the acoustic waveform, and these features were later combined to estimate the subjective ratings of TE speech. This approach was validated with a database of subjective ratings of speech samples recorded from 35 TE speakers. Results showed moderate correlations between the objective metrics and the subjective ratings, and these correlations were significantly better than those obtained with traditional methods used in the telecommunication applications.
全喉切除术通常是许多诊断出患有晚期喉癌的患者的首选治疗方法。该手术改变了正常的语音产生机制，气管食管（TE）语音是喉切除术后发声的一种替代方法。TE 语音是通过肺部空气穿过上食管产生振动源，然后将其发成语音而创建的。TE 语音通常具有异常的音质。对 TE 语音的声学分析具有量化语音质量的潜力，并有助于言语语言病理学家进行康复。受电信行业在语音质量估计方面的最新进展的启发，本文研究了听觉模型在预测正常听力听众对 TE 语音的评分中的应用。采用 Moore-Glasberg 听觉模型从声谱中提取感知相关的特征，然后将这些特征组合起来估计 TE 语音的主观评分。该方法使用从 35 位 TE 说话者录制的语音样本的主观评分数据库进行了验证。结果表明，客观指标与主观评分之间存在中等相关性，并且这些相关性明显优于电信应用中使用的传统方法获得的相关性。
