Quantitative measurement of speech sound distortions due to inadequate dental mounting

Abstract

In this paper, the most adequate quantitative parameters are sought that reflect the dependence of speech sound distortions on the correct position of dental mountings. We suggest the Hurst fractal exponent and some parameters calculated from the Paul wavelet transform of speech sound are useful as quantitative parameters. The investigations are focussed on the alterations of the first /s/ phoneme when the patient produces the Sisyphe sound. The results will be useful to obtain a rigorous, computer assisted method to establish (design) the optimal parameters of prosthetic mounts in order to ensure best speech quality.

Quantitative measurement of speech sound distortions with the aid of minimum variance spectral estimation method for dentistry use

“…We must also underline that the MVSEM introduced in this paper gives clear and expressive information about the speech sounds quality and the speech sounds alteration, as by the Fourier transform, the wavelet transform and the spectrogram analysis, used in the previous papers (Drȃgȃnescu et al 2003(Drȃgȃnescu et al , 2006Sinescu et al 2008). This is also evident if Figure 3 is compared with Figures 2 and 4.…”

“…In these papers (Drȃgȃnescu et al 2003(Drȃgȃnescu et al , 2006Sinescu et al 2008), it can be seen that the Fourier transform, the wavelet transform and the spectrogram give complex forms. These forms are difficult to interpret and estimate quantitatively.…”

“…The aim of this paper is to continue the investigations done in our previous papers (Drȃgȃnescu et al 2003(Drȃgȃnescu et al , 2006Sinescu et al 2008) to establish some quantitative parameters from the analysis of the speech sound to build a sensitive method to establish the correct placement of the dental prostheses, which sensitively express the minimal distortion of speech sounds. In these papers, quantitative parameters have been introduced that reflect, in an accurate manner, invariant personal characteristics of the human voice extracted from power spectra (Drȃgȃ-nescu et al 2003), fractal Hurst exponent and wavelet spectrum (Drȃgȃnescu et al 2006), Shannon's entropy, wavelet and time -frequency analysis (Sinescu et al 2008).…”

“…In these papers, quantitative parameters have been introduced that reflect, in an accurate manner, invariant personal characteristics of the human voice extracted from power spectra (Drȃgȃ-nescu et al 2003), fractal Hurst exponent and wavelet spectrum (Drȃgȃnescu et al 2006), Shannon's entropy, wavelet and time -frequency analysis (Sinescu et al 2008).…”

Fractals in dentistry

“…Drȃ gȃ nescu et al [34] performed measurements of speech distortion after prostheses positioning in two Class IV Kennedy patients. They used a microphone to record speech sound signals and a professional sound analysis software to process them.…”