Elsevier

Journal of Voice

Volume 24, Issue 6, November 2010, Pages 651-653
Journal of Voice

Cepstral Analysis of Voice in Persons With Vocal Nodules

https://doi.org/10.1016/j.jvoice.2009.07.008Get rights and content

Summary

Objectives

Traditional measures of dysphonia were reported to be unreliable, and the measurements of cepstral peak prominence (CPP) have been shown to correlate well with perceptions of breathiness. Hence, the present study was hypothesized that there would be abnormal reduction of the cepstral peak in the speakers with vocal nodules relative to the normal controls as they are characterized by hoarse and breathy voice due to inadequate closure of vocal folds.

Method

Phonation of sustained vowel /a/ was subjected to acoustic analysis using Computerized Speech Lab (CSL model 4150) software. The vowels were analyzed acoustically with the measurement of CPP.

Results

Independent t test was employed to compare the significant differences between the clinical and the control groups in both males and females. The results revealed significant differences across the two groups at P < 0.05. The obtained results were discussed with respect to the underlying pathophysiology.

Conclusion

The present study investigated the CPP in subjects with vocal nodules. Cepstral deviations in the clinical group are explained due to the presence of laryngeal pathology leading to the lower values of CPP.

Introduction

Voice is a multidimensional series of measurable events. It is defined as the laryngeal modulation of the pulmonary airstream, which is further modified by the configuration of the vocal tract.1 The movements of the folds are controlled by the biomechanical properties of the folds themselves, magnitude of the air pressure beneath the folds, and their neural control.2 Voice clinicians do require an objective, reliable, and relatively automatic method to assess voice change after medical, surgical, or behavioral intervention. Although acoustic signs are the imperfect mirrors of underlying vocal fold physiology, there is a great deal of correspondence between the vocal fold physiology and the voice acoustics, and much can be inferred about the physiology based on the acoustics. These acoustic measures have been developed to quantify and differentiate the vocal parameters in both normal and dysphonic voices. A normal voice signal is said to be periodic and should have very little cycle-to-cycle variability in frequency and amplitude. There are several algorithms used to calculate these periodicity in the normal as well as the dysphonic population. Despite their ability to quantify the acoustic measures, traditional measures such as noise-to-harmonic ratio (NHR), jitter, and shimmer are shown to be the unreliable predictors of dysphonia in many studies.3 One reason could be that the dysphonia may not be truly associated with any perturbation measure. Another possibility could be that these measurements rely on the ability to calculate the correct F0, small errors in determining F0 can result in significant errors in measuring perturbations and NHR.4 Because of the difficulty in determining F0, accurate measures of periodicity in dysphonic voice are difficult to obtain.

Measures of cepstrum attempt to resolve this problem. Cepstrum is a Fourier transformation of a spectrum.5, 6 To produce a cepstrum, Fourier transformation of an acoustic signal is first performed to create a spectrum, and hence, the voice signal is transformed from the time domain to the frequency domain. Thus, the intensity of each frequency within the signal is shown in the spectrum. Performing a Fourier transformation of the spectrum then produces the cepstrum. In doing so, the signal is transformed from the frequency domain to the quefrency that equals 1/frequency. This gives a better picture of the degree of harmonic organization. Unlike the measures of frequency and amplitude perturbation measure, this measure does not rely on the accurate calculation of fundamental frequency (F0). Rather, they rely on the peak-to-average calculation. Because of this reason, cepstral measures tend to be more reliable than other measures of periodicity.4

A normal voice having well-defined harmonic structure would have a strong cepstral peak compared with the breathy and hoarse voice as there is a poorly defined harmonic structure evidenced in breathy and hoarse voices. Cepstral peak prominence (CPP) and smoothed CPP are the best predictors of overall dysphonia in comparison to NHR, amplitude perturbation quotient, relative average perturbation, and smoothened pitch perturbation quotient.3, 7, 8, 9 They reported that none of these measures other than cepstral measure were useful in predicting dysphonia.

With the present development of digital technology and methods for processing speech, it is possible to make effective objective acoustic diagnostics for assessment and management of voice disorders. These methods are useful as all pathologies and diseases of the human vocal tract influence the quality of a patient's speech signal. The studies reviewed above suggest the potential clinical applications of cepstral analysis measure. However, in the contemporary clinical outlook, its popularity is limited. In the present study, it was hypothesized that there would be abnormal reduction of the cepstral peak in the speakers with vocal nodules compared with the normal controls as they are characterized by hoarse and breathy voice due to inadequate closure of vocal folds. With this hypothesis, the present study was carried out with an aim of examining the cepstral measure in persons with vocal nodules and compare it with the normal controls.

The aim of the study is to measure CPP in subjects with vocal nodules.

Section snippets

Method

The study followed a prospective case-control design.

Results

The present study investigated the CPP in patients with vocal nodules. Results are shown in Table 1.

The results in Table 1 reveal lower values of CPP in speakers with vocal nodules. The results revealed significant differences across the two groups in both males and females at P < 0.05.

Discussion

Digital signal processing techniques have been used to perform an acoustic analysis for vocal quality assessment. This is due to the simplicity and the noninvasive nature of the measurement procedure. This is valuable in the diagnosis of pathological voices and may be used as complementary tool in laryngological evaluation. Cepstral-based estimation is one such measure that is used to provide a baseline estimate of the noise level in the logarithmic spectrum for speech. This gives a better

Conclusions

The present study investigated the CPP in the speakers with vocal nodules, and the results reveal lower values of cepstral peak in the clinical group in comparison with the control group. Cepstral deviations in the clinical group are explained due to the presence of benign localized lesion in the vocal folds, leading to the lower values of CPP in the speakers with vocal nodules. The results presented here suggest the applicability of CPP in the clinical practice of otolaryngologist and the

References (9)

There are more references available in the full text version of this article.

Cited by (63)

View all citing articles on Scopus
View full text