CPPS distributional shape and parameters as effective tools to discriminate dysphonic and healthy voice

Objective. Voice is the main professional tool for many people, therefore it is important to preserve it and to prevent damages at the vocal apparatus. Voice monitoring and objective measures of the vocal behavior and vocal health can warn a talker against risky situations and highlight existing problems at the vocal apparatus. Previous works have investigated the parameter Cepstral Peak Prominence Smoothed (CPPS) as a possible indicator of vocal health status, giving its mean value and standard deviation as promising results. The objective of this work is to determine whether the CPPS distributional shape and its parameters can discriminate dysphonic and healthy voice. Methods. Authors are building a well-assessed vocal database in order to validate the proposed parameters. Vocal data is collected at the San Giovanni Battista hospital in Turin. First, an otolaryngologist performs the laringostroboscopy exam and the perceptual assessment of voice (GIRBAS scale) to voluntary patients. Then, three different speech materials (vocalization, reading and free-speech) are simultaneously acquired at a microphone in air, placed at a fixed distance from the mouth, and at a contact microphone attached at the jugular notch. Last, each patient is asked to fill in a questionnaire about its perception of voice in daily life (PAPV). Results. At the moment, 25 patients make up our study group and 8 subjects belong to the control group. Recruitment will end in November. Associations between the presence/absence of any type of dysphonia, the values of the CPPS parameters, the perceptual evaluation of voice and the disability level felt in everyday life due to vocal problems will be investigated. Furthermore, new techniques of classification and data mining based on CPPS distributional shape will be performed Conclusions. The proposed approach can have implications both on clinical practice and on the prevention of vocal problems, especially for voice professionals.