Multiparametric quantitative evaluation of the speech production system: a study in normal subjects

This paper describes a systems architecture developed for the study of the speech production system. The architecture utilizes two personal computers: the "Word-Image Presenter" (WIP) that presents a sequence of words to the subject under examination as visual stimulation, and the "Signal Acquisition System" that controls the WIP, acquires the acoustic signal, the oromandibular electromyographic signals, the kinematic lip and jaw signals and measures reaction times and durations. Thirty normal subjects divided into three age groups (18-44, 45-59, 60-80) underwent examination by means of the described system. During the test a random sequence of two words was presented to the subjects; the protocol consisted of an immediate reading task and two delayed reading tasks with variable foreperiods of 0.1 and 1.5 s for one task and 0.5 and 4 s for the other. The analysis of the resultant acoustic and electromyographic signals indicated that reaction times increased with age in both the immediate and delayed tasks. The analysis of variance showed that the difference between the young and elderly groups was statistically significant. The sample size of the groups, however, is too small to consider these results normative data.