Jan Vokral

Computational modelling of effect of tonsillectomy on voice production

The aim of this study is to develop mathematical 3D finite element (FE) models for numerical simulations of vibroacoustic properties of the human vocal tract after a tonsillectomy. Similar experimental studies are not easily realizable on living subjects. The FE models of the acoustic spaces corresponding to the human vocal tract for the Czech vowels /a:/ and /i:/ and the acoustic space around the human head were used in numerical simulations of phonation. The acoustic resonant characteristics of the FE models were studied using modal and transient analyses (excitation by a short pulse). Calculated results show that a tonsillectomy causes a frequency shift of the 3rd (down by ∼180 Hz) and 4th (down by ∼120 Hz) formants down to the lower frequencies for the vowel /a:/ and similarly for the 2nd, 4th and 5th formants for the vowel /i:/ (all down by ∼100 Hz). Similar effects and results can be found in experimental studies in literature. The formant changes are dependent on the size of additional acoustic spaces that occur after a tonsillectomy. The verification of the model was performed on the recordings of patients before and after the tonsillectomy operation. Multi-Dimensional Voice Program (MDVP Advanced, KAY Elemetrics Corp.) was used for the comparing of the formant centre frequencies. Very small differences in the results of subjective evaluation of the voice before and after tonsillectomy were found.

Acoustic Parameters for the Evaluation of Voice of Future Voice Professionals

The authors tried to establish quantitative and qualitative acoustic parameters of a good voice, suitable for future voice professionals. In their work they used long-time average spectrum analysis (LTAS) and three-dimensional analysis of periodicity (3D-PAN). They consider the regression straight line of formant regions and the parameters offered by 3D-PAN--jitter first of all--as the main acoustic parameters for the evaluation of voice quality and draw attention to the fact that acoustic parameters represent only one part of the evaluation of voice quality.

Bayesian changepoint detection for the automatic assessment of fluency and articulatory disorders

The accurate changepoint detection of different signal segments is a frequent challenge in a wide range of applications. With regard to speech utterances, the changepoints are related to significant spectral changes, mostly represented by the borders between two phonemes. The main aim of this study is to design a novel Bayesian autoregressive changepoint detector (BACD) and test its feasibility in the evaluation of fluency and articulatory disorders. The originality of the proposed method consists in its normalizing of a posteriori probability using Bayesian evidence and designing a recursive algorithm for reliable practice. For further evaluation of the BACD, we used data from (a) 118 people with various severity of stuttering to assess the extent of speech disfluency using a short reading passage, and (b) 24 patients with early Parkinsons disease and 22 healthy speakers for evaluation of articulation accuracy using fast syllable repetition. Subsequently, we designed two measures for each type of disorder. While speech disfluency has been related to greater distances between spectral changes, inaccurate dysarthric articulation has instead been associated with lower spectral changes. These findings have been confirmed by statistically significant differences, which were achieved in separating several degrees of disfluency and distinguishing healthy from parkinsonian speakers. In addition, a significant correlation was found between the automatic assessment of speech fluency and the judgment of human experts. In conclusion, the method proposed provides a cost-effective, easily applicable and freely available evaluation of speech disorders, as well as other areas requiring reliable techniques for changepoint detection. In a more modest scope, BACD may be used in diagnosis of disease severity, monitoring treatment, and support for therapist evaluation.

Emotions in the Sight of Long-Time Averaged Spectrum and Three-Dimensional Analysis of Periodicity

The authors used long-time averaged spectrum and three-dimensional analysis of periodicity (3D-PAN) for analyzing emotional expressions in utterances. They confirmed the movement of the skewness of formant region, that is getting smaller at anger and at joy. It can reach positive values. The skewness is higher at sadness. These findings correspond to the physiological phenomena that we can follow in the whole musculature. The 3D-PAN has not brought any new finding as far.

Finite element model of supraglottal space in cleft palate

Abstract An approximate finite element (FE) model of acoustic spaces corresponding to the human adult male vocal tract and nasal cavity was created according to the geometrical data published in literature. The FE model for the vowel / a / was analyzed for increasing size of a cleft, which joints both acoustic spaces. The influence of the cleft palate on the pronunciation of the vowel / a / is studied using a modal analysis of the FE model. The results of numerical solutions are in reasonable agreement with clinical observations.