Electroglottographic and Acoustic Parametrization of Phonatory Quality Provide Voice Profiles of Pathological Speakers

OBJECTIVES

The present study firstly tries to find subgroups of pathological male and female phonation using data from a number of 534 pathological speakers. Secondly, this subgroup classification provides a basis for achieving voice profiles of pathological phonatory quality.

METHODS

Using complementarily orientated electroglottographic and acoustic parametrization of phonatory quality, sustained vowel productions of 267 male and 267 female speakers were considered.

RESULTS

In a first step, a clustering technique differentiates three subgroups within each gender on the basis of the EGG- and three subgroups on the basis of the acoustic parameters. In a second step, this subgroup definition allows one to present voice profiles of pathological speakers by combining the parameter means of the electroglottographically determined subgroups with those of the acoustically determined subgroups.

CONCLUSIONS

The presented voice profiles provide a finer reference basis for the classification of different pathological phonation types as well as for the evaluation of shifts in individual phonatory behavior due to therapy or spontaneous recovery.

fMRI and acoustic analyses reveal neural correlates of gestural complexity and articulatory effort within bilateral inferior frontal gyrus during speech production

In an event-related fMRI study of overt speech production, we investigated the relationship between gestural complexity and underlying brain activity within bilateral inferior frontal gyrus (IFG). We operationalized gestural complexity as the number of active articulatory tiers (glottal, oral, nasal) and the degree of fine-grained temporal coordination between tiers (low, high). Forty-three neurotypical participants produced three types of highly-frequent non-word CV-syllable sequences, which differ systematically in gestural complexity (simple: ['dadada], intermediate: ['tatata], complex: ['nanana]). Comparing blood oxygen level-dependent (BOLD) responses across complexity conditions revealed that syllables with greater gestural complexity elicited increased activation patterns. Moreover, when durational parameters were included as covariates in the analyses, significant effects of articulatory effort were found over and above the effects of complexity. The results suggest that these differences in BOLD-response reflect the differential contribution of articulatory mechanisms that are required to produce phonologically distinct speech sounds.

Evaluation of Phonatory Behavior and Voice Quality in Patients with Multiple Sclerosis Treated with Deep Brain Stimulation

OBJECTIVES

The effect of deep brain stimulation (DBS) on phonatory behavior and voice quality in eight patients with multiple sclerosis (MS) was examined instrumentally and perceptually. The acoustic signals of vowel productions obtained from patients (produced with and without stimulation) and from a group of 16 healthy control speakers were analyzed to prove statistically the changes of phonatory behavior and voice quality.

STUDY DESIGN

This is a randomized study.

METHODS

Firstly, a new parametrization was used to determine phonatory behavior. Secondly, a perceptual evaluation of voice quality of the same speech material was performed.

RESULTS

With stimulation, phonation has a greater tendency to be strained. The results of perceptual evaluation support this strained phonation behavior under stimulation, resulting in a smaller degree of breathiness ratings of all raters. Without stimulation, an impaired and partly disturbed adduction of the vocal folds can be shown. These findings are also supported in the perceptual experiment providing a higher degree of hoarseness ratings of all raters for these signals.

CONCLUSIONS

High-frequency electrical impulses to the thalamus in patients with MS influence the phonatory behavior of their vocal folds. The results suggest the need for long-term monitoring of phonatory behavior during DBS to initiate adequate treatments without delay.

Improvement of upper limb ataxia and intention tremor allowing cessation of thalamic electrostimulation after four years

A 34-year-old female patient suffering from multiple sclerosis was treated with thalamic electrostimulation for right dominant brachial ataxia and intention tremor. At the end of the fourth year, the energy of the battery of the impulse generator was depleted. However, the patient was able to use her hands without major impairment with the stimulator OFF, due to a stable reduction of the degree of ataxia and intention tremor. The opposite thalamus received an electrode that was never activated because of a permanent thalamotomy effect. Thus, it can be concluded that restoration of function by means of deep brain stimulation might have been beneficial towards achieving a reorganization and stabilization of subcortico-cortical and cerebellar circuitry, supporting the process of self-repair in this patient with a less aggressive course of multiple sclerosis. However, scientific proof has yet to be found. This exceptional observation emphasizes the need for longterm studies.

Stimmprofile zur Normalstimme auf der Grundlage akustischer und elektroglottografischer Analysen

Hintergrund: Die vorliegende Studie beschäftigt sich mit der Dimensionierung normalstimmlicher Phonation. Es wird allgemein akzeptiert, dass diese Dimensionierung mit einer großen Bandbreite einhergeht.

Material und Methoden: Unter komplementärer Verwendung akustischer und elektroglottografischer Analysemethoden zur Parametrisierung von Stimmqualität werden gehaltene Vokalproduktionen von 267 männlichen und 267 weiblichen Sprechern analysiert. Clusterverfahren ­dienen auf der Grundlage dieser Analyseergebnisse zur Definition geschlechtsspezifischer ­Subgruppen.

Ergebnisse: Zunächst werden für beide Geschlechter getrennt sowohl auf der Basis der elektroglottografischen als auch der akustischen Analyse 3 stabile Subgruppen geliefert. Diese Subgruppendefinition ermöglicht darüber ­hinaus durch eine Kombination der Parametermittelwerte beider Analyseverfahren die Darstellung von Stimmprofilen.

Schlussfolgerungen: Die präsentierten Stimmprofile sollen als Referenzstimmprofile eine solidere Basis zur Klassifikation verschiedener normalstimmlicher und pathologischer Phonationstypen liefern und darüber hinaus dazu dienen, Veränderungen des individuellen Phonationsverhaltens aufgrund von therapeutischen Ansätzen oder spontaner Genesung evaluieren zu können.

[Functional MRI of speech]

Speech is a multifactorial term. In relationship with our central organ, speech is a communication medium which in a complex manner includes the production and perception of verbal information. Even within this limitation, imaging exploration of the neuronal basis of verbal communication offers an extensive investigative field which can only be presented in this article as highlights. The main focus is on the technical hurdles of investigations into speech production using functional magnetic resonance imaging (fMRI) and solution options will be presented. The difficulties contained in this aspect are that spoken words can only be poorly assessed in the noise level of the measurements and that the movements of the mouth necessary for speech motoric can lead to technical imaging artefacts. Investigation techniques and designs will be discussed which allow a representation of speech-relevant brain centers despite these problems.

Effect of bilateral stimulation of the subthalamic nucleus on different speech subsystems in patients with Parkinson's disease

The effect of deep brain stimulation on the two speech-production subsystems, articulation and phonation, of nine Parkinsonian patients is examined. Production parameters (stop closure voicing; stop closure, VOT, vowel) in fast syllable-repetitions were defined and measured and quantitative, objective metrics of vocal fold function were obtained during vowel production. Speech material was recorded for patients (with and without stimulation) and for a reference group of healthy control speakers. With stimulation, precision of the glottal and supraglottal articulation as well as the phonatory function is reduced for some individuals, whereas for other individuals an improvement is observed. Importantly, the improvement or deterioration is determined not only on the basis of the direction of parameter change but also on the individuals' position relative to the healthy control data. This study also notes differences within an individual in the effects of stimulation on the two speech subsystems. These findings qualify the value of global statements about the effect of neurostimulatory operations on Parkinsonian patients. They also underline the importance of careful consideration of individual differences in the effect of deep brain stimulation on different speech subsystems.

Effect of Deep Brain Stimulation on Different Speech Subsystems in Patients with Multiple Sclerosis

The effect of deep brain stimulation on articulation and phonation subsystems in seven patients with multiple sclerosis (MS) was examined. Production parameters in fast syllable-repetitions were defined and measured, and the phonation quality during vowel productions was analyzed. Speech material was recorded for patients (with and without stimulation) and for a group of healthy control speakers. With stimulation, the precision of glottal and supraglottal articulatory gestures is reduced, whereas phonation has a greater tendency to be hyperfunctional in comparison with the healthy control data. Different effects on the two speech subsystems are induced by electrical stimulation of the thalamus in patients with MS.

Multiparametrische Stimmprofildifferenzierung zu männlichen und weiblichen Normalstimmen auf der Grundlage akustischer Analysen

BACKGROUND

An improved method to obtain voice quality parameters from the acoustic speech signal is presented and applied to the determination of voice profiles.

METHODS

The method is based on voice parameters proposed by Stevens and Hanson (1994). Vowel productions of 300 adult speakers (150 male and 150 female speakers) without any known voice pathologies were considered.

RESULTS

In a first step, gender differentiated mean normal voice profiles are presented. The significant differences between male and female speakers indicate the necessity for separate treatment of the two groups. In a second step, a clustering technique differentiates three numerically substantial subgroups for male and female speakers. Over and above that, the acoustic parameters with the best potential for subgroup differentiation are determined. Additionally, a continuum containing individual variation within normal phonation is shown.

CONCLUSIONS

The relevance of the analysis method is supported by the possibility of a physiological interpretation of the voice data.

Correlates of Varying Vocal Fold Adduction Deficiencies in Perception and Production: Methodological and Practical Considerations

In this study the voice characteristics of normal male and female speakers are compared to those of two groups of patients with unilateral vocal fold paralysis. In order to enhance phonation, the patients in the first group compensate for the adduction deficiency which results from paralysis. The patients in the second group do not use compensatory strategies. Sustained vowels [i:, a:, u:] were produced by the speakers and scored for roughness, breathiness and hoarseness (RBH) by 8 raters. Although interrater agreement for RBH scores is only moderate on average, these percepts make consistent distinctions between the three speaker groups. Consistent but different distinctions are made between the three speaker groups for male and female speakers. The results show that male and female speakers should not be pooled in experimental studies of the pathological voice. Our results also indicate that female patients with a compensated unilateral vocal fold paralysis cannot be clinically evaluated solely on the basis of perception, because their voices cannot be distinguished from normal, healthy female speakers, despite their physiological impairment. The group distinctions made on the basis of RBH scores are supported by differences in the acoustic parameters which are derived by automatic analysis of the sustained vowels. Despite identical group distinctions for RBH scores and acoustic parameters, the acoustic basis of the percepts is not straightforward. Different acoustic predictors of the percepts were found for male compared to female speakers. Additionally, interrater differences point towards the presence of perceptual trading relations.

Bilateral high-frequency electrical impulses to the thalamus reduce voice tremor: acoustic and electroglottographic analysis. A case report

An 81-year-old female patient suffering from disabling Holmes' tremor affecting both upper extremities, the head and additionally the vocal apparatus underwent bilateral thalamic ventralis intermedius nucleus (v.i.m.) stimulation. With the stimulation ON, the patient experienced complete suppression of the limb and head tremor and thorough voice normalization. Acoustic and electroglottographic (EGG) analysis showed a tendency towards hyperfunctional phonation with the stimulation ON as well as OFF, but a less disturbed vocal cord vibration pattern with the stimulation ON in comparison with a group of normal female speakers. This example shows that long-term monitoring of the vocal apparatus under deep brain stimulation therapy (DBS) of movement disorders must be planned in order to modify the stimulation parameters, if necessary, or to initiate logopaedic treatment.

[Effect of deep brain stimulation on glottal phonation in patients with Parkinson's disease and multiple sclerosis]

The present study examines the effect of neurostimulatory operations on glottal phonation of 3 parkinsonian patients and 3 patients with multiple sclerosis. With the help of two voice analysis programs (MDVP from Kay Elemetrics and EEG Program by Marasek) for the acoustic and electroglottographic definition of voice characteristics, vowel productions of the patients, which were recorded under two conditions (with and without stimulation), were analysed. In a first step, significantly different intrasubject means in the two conditions indicate the effect of neurostimulation. The strength of the effect differs among subjects, particularly in the case of patients with Parkinson's disease. In a second step, a gender-differentiated comparison of the individual patient's data (recorded with and without stimulation) with a group of normal voice speakers (150 male and 150 female speakers) is carried out. This intersubject comparison proves useful in that it relativizes the results from the intrasubject comparison. It is shown for the parkinsonian patients that stimulation causes a relative deterioration of the glottal cycle, while for the patients with multiple sclerosis a tendency for hyperfunctional phonation is observed. In the latter case, the results suggest the need for long-term monitoring of phonation behaviour during chronic electrical stimulation.

[Multiparametric description of voice quality for normal male and female voices]

The present study examines the relevance of the voice parameters of two commercially available voice analysis programs (MDVP and EGG Program from Kay Elemetrics) for the acoustic and electroglottographic definition of normal voice profiles for men and women. Vowel productions of 300 adult speakers (150 male and 150 female speakers) without any known voice pathology were analyzed. In a first step, gender-differentiated mean normal voice profiles are presented for each of the two analysis programs. The significant differences between the male and female speakers indicate the necessity for the separate treatment of the two groups. In a second step, a clustering technique differentiates two numerically substantial subgroups within each gender on the basis of the MDVP parameters. The EGG parameters do not achieve the same degree of subgroup differentiation. In a factor analysis, the relative weighting of the MDVP parameters within the factors also indicates their better suitability for normal voice differentiation than the EGG parameters as they are defined at present.