Complementary approaches for voice disorder assessment

International audienceDysphonia is a symptom of laryngeal failure in producing periodic sound. At the opposite of dysarthria which is a neurologic speech disorder which often induces a problem of intelligibility, dysphonia is often considered as a "minor" trouble linked to an esthetic point of view. An explanation of this sub-consideration is that voice quality is generally described as a paralinguistic phenomenon with little impact on communication. However, the voice remains the support of the spoken communication: what becomes speech without voice?Moreover, the social relevance and economical impact of voice disorders is now obvious, especially for school teachers which represent, in occidental countries, the largest group of professionals who use their voice as a primary tool of trade. For instance, a recent study conducted in Belgium on 723 school teachers has revealed that 10.5% of them are clearly suffering of voice disorder, this ratio being more important among young and inexperienced teachers.Instrumental assessment of voice quality has been explored these twenty last years especially at the request of clinicians ENT, phoniatricians and speech therapists. The goal is to obtain a quantitative report of the voice quality for the assistance with the diagnosis and the follow-up of pathology. In the instrumental approaches containing acoustic measurements, aerodynamic or physiological like those carried out with EVA equipment, the evaluation is clearly turned towards a determination of the "mechanical" characteristics of the phonatory system : the speaker produces sustained vowel, sequences of ,/papapa /, glissandos, maximum phonation time … On the other hand, the perceptual evaluation, where a panel of listeners evaluate subjectively the voice quality on continuous speech, is closed to spoken communication. However, if these two procedures of assessment are carried out in parallel on the same patients, we can obtain an agreement of these two evaluations in more than 80% of cases. That means that the evaluation of the biomechanics, the motor control, the stability and the regulation of the pneumo phonatory system is essential in the study of dysphonia.We will present these comparative studies of voice quality assessment based on complementary approaches. The first study was undertaken on 449 speakers whose voice quality was evaluated in parallel by a perceptual judgment and objective measurements on acoustic and aerodynamic data. Results showed that a non-linear combination of 7 parameters allowed the classification of 82% voice samples in the same grade as the jury. The second study relates to the adaptation of Automatic Speaker Recognition (ASR) techniques to pathological voice assessment. Experiments conducted on 80 female voices provide promising results, underlining the interest of such an approach. We benefit from the multiplicity of theses techniques to evaluate the methodological situation which points fundamental differences between these complementary approaches (bottom-up vs top-down, global vs analytic). We also discuss some theoretical aspects about relationship between acoustic measurement and perceptual mechanisms which are often forgotten in the performance race