Influence of Botox on adductor spasmodic dysphonic speech: Perceptual-acoustic analyses of voice quality

By M.P. Cannito, B.K. Bender, R.A. Dressler, E.H. Buder, G.E. Woodson, and T. Murry

Acoustic variables, including standard deviation of fundamental frequency, mean fundamental frequency, standard deviation of amplitude, and signal-to-noise ratio, were extracted via digital signal analysis (Cspeech) from oral readings of the first paragraph of The Rainbow Passage. Speech was recorded prior to and approximately one month following initial botulinum toxin injection of the vocal folds in 10 speakers with adductor spasmodic dysphonia and in 10 matched normal controls. The same speech samples were subjected to perceptual scaling by six expert voice disorders clinicians under rigorously controlled psychoacoustic laboratory conditions. Results indicated that perceived voice quality as well as variability of fundamental frequency improved significantly following Botox, but remained significantly poorer than normal levels. Mean fundamental frequency and amplitude variability did not change. Scaling judgments of voice quality were significantly correlated with variability of fundamental frequency, variability of amplitude, and signal-to-noise ratio. Multiple regression showed that 77% of the variance in perceptual scaling judgments could be explained by a linear combination of the four acoustic variables. These findings suggest that while speech improved in response to treatment it remained significantly abnormal, not all acoustic variables improved, and that listener responses were closely tied to specific acoustic characteristics of the connected speech signal.