A STUDY OF MODELING PATHOLOGICAL VOCAL CORDS WITH APPLICATION TO THE DIAGNOSIS OF LARYNGEAL DISEASES THROUGH THE ACOUSTIC ANALYSIS OF HOARSE VOICE

点击翻译按钮获取中文摘要

(1) We have developed several novel models of pathological vocal cords including vocal cords with nodules or a single polyp using the acoustic analysis of hoarse voices and video pictures of pathological vocal cords with relevant diagnoses made by otolaryngologists.(2) We have investigated relations between physical conditions of those models and acoustic features of hoarse voice synthesized with the models.(3) We have found some acoustic features of hoarse voice, in terms of which it is possible to effectively represent hoarse voices. They are found to be useful for the diagnosis and estimation of some physical features of pathological vocal cords. This finding has enabled us to develop a new noninvasive procedure for estimating vocal cord nodule or polyp features such as the mass and stiffness of nodule or polyp through the acoustic analysis of hoarse voice.Findings from the study are as follows :[1] A novel model of pathological vocal cords is able to simulate the subtle movement of the vocal cords in the presence of aspiration noise in the glottis and is able to produce synthetic voices which mimic patients' voices.[2] By the newly developed procedure it is possible to estimate dimensions of vocal cord nodules only through hoarse voice analysis.[3] Estimated noise power spectrum and waveform information such as amplitudes and phases of harmonic frequency components of hoarse voice are useful for estimating vocal cord nodule features.