Investigation on teachers’ vocal behavior during a school year using smartphone devices

The vocal-behaviour assessment of occupational voice users was a tricky task when devices able to collect vocal parameters during daily working activities were unavailable. A first attempt to overcome this problem was made with the development of wearable vocal analysers equipped with contact sensors, but these devices only provide vocal-load related parameters (voice intensity, pitch and vocal doses) and their high cost prevents from monitoring a large number of subjects. A recently proposed solution to overcome such limitations, which has been also adopted in this work, consists in using smartphone devices to collect voice data. This paper deals with the results obtained during an experimental campaign that took place in kindergarten and primary schools and that involved healthy teachers. Data collection and processing were performed by means of a smartphone-based platform that includes: a piezoelectric contact microphone that senses the vibration at the base of the neck of the subject under monitoring; a smartphone device that acquires and pre-processes the signal at the output of the contact microphone; specifically designed MatLab scripts that estimate parameters related not only to vocal loading but also to voice quality, such as statistics of Cepstral Peak Prominence Smoothed (CPPS) distributions. The 14 subjects that participated in the experiment followed the same protocol at the beginning and at the end of the same school year: emission of a sustained vowel, free speech at comfortable level (5 min), monitoring during lessons (4 hours). The parameters estimated during the two periods were very similar: jitter, shimmer and standard deviation of CPPS distribution for the sustained vowel confirmed the healthy state of the subjects, while loading parameters (difference in voice intensity and phonation time with respect to the free speech) and quality parameters (95° percentile of CPPS distribution) during teaching activities highlighted possible risks for vocal health.