Spatiotemporal classification of vocal fold dynamics by a multimass model comprising time-dependent parameters

A model-based approach is proposed to objectively measure and classify vocal fold vibrations by left-right asymmetries along the anterior-posterior direction, especially in the case of nonstationary phonation. For this purpose, vocal fold dynamics are recorded in real time with a digital high-speed camera during phonation of sustained vowels as well as pitch raises. The dynamics of a multimass model with time-dependent parameters are matched to vocal fold vibrations extracted at dorsal, medial, and ventral positions by an automatic optimization procedure. The block-based optimization accounts for nonstationary vibrations and compares the vocal fold and model dynamics by wavelet coefficients. The optimization is verified with synthetically generated data sets and is applied to 40 clinical high-speed recordings comprising normal and pathological voice subjects. The resulting model parameters allow an intuitive visual assessment of vocal fold instabilities within an asymmetry diagram and are applicable to an objective quantification of asymmetries.     

Computation of physiological human vocal fold parameters by mathematical optimization of a biomechanical model

With the use of an endoscopic, high-speed camera, vocal fold dynamics may be observed clinically during phonation. However, observation and subjective judgment alone may be insufficient for clinical diagnosis and documentation of improved vocal function, especially when the laryngeal disease lacks any clear morphological presentation. In this study, biomechanical parameters of the vocal folds are computed by adjusting the corresponding parameters of a three-dimensional model until the dynamics of both systems are similar. First, a mathematical optimization method is presented. Next, model parameters (such as pressure, tension and masses) are adjusted to reproduce vocal fold dynamics, and the deduced parameters are physiologically interpreted. Various combinations of global and local optimization techniques are attempted. Evaluation of the optimization procedure is performed using 50 synthetically generated data sets. The results show sufficient reliability, including 0.07 normalized error, 96% correlation, and 91% accuracy. The technique is also demonstrated on data from human hemilarynx experiments, in which a low normalized error (0.16) and high correlation (84%) values were achieved. In the future, this technique may be applied to clinical high-speed images, yielding objective measures with which to document improved vocal function of patients with voice disorders.     

Model-based classification of nonstationary vocal fold vibrations

Classification of vocal fold vibrations is an essential task of the objective assessment of voice disorders. For historical reasons, the conventional clinical examination of vocal fold vibrations is done during stationary, sustained phonation. However, the conclusions drawn from a stationary phonation are restricted to the observed steady-state vocal fold vibrations and cannot be generalized to voice mechanisms during running speech. This study addresses the approach of classifying real-time recordings of vocal fold oscillations during a nonstationary phonation paradigm in the form of a pitch raise. The classification is based on asymmetry measures derived from a time-dependent biomechanical two-mass model of the vocal folds which is adapted to observed vocal fold motion curves with an optimization procedure. After verification of the algorithm performance the method was applied to clinical problems. Recordings of ten subjects with normal voice and ten dysphonic subjects have been evaluated during stationary as well as nonstationary phonation. In the case of nonstationary phonation the model-based classification into "normal" and "dysphonic" succeeds in all cases, while it fails in the case of sustained phonation. The nonstationary vocal fold vibrations contain additional information about vocal fold irregularities, which are needed for an objective interpretation and classification of voice disorders.     

Three-dimensional biomechanical properties of human vocal folds: parameter optimization of a numerical model to match in vitro dynamics

The human voice signal originates from the vibrations of the two vocal folds within the larynx. The interactions of several intrinsic laryngeal muscles adduct and shape the vocal folds to facilitate vibration in response to airflow. Three-dimensional vocal fold dynamics are extracted from in vitro hemilarynx experiments and fitted by a numerical three-dimensional-multi-mass-model (3DM) using an optimization procedure. In this work, the 3DM dynamics are optimized over 24 experimental data sets to estimate biomechanical vocal fold properties during phonation. Accuracy of the optimization is verified by low normalized error (0.13 ± 0.02), high correlation (83% ± 2%), and reproducible subglottal pressure values. The optimized, 3DM parameters yielded biomechanical variations in tissue properties along the vocal fold surface, including variations in both the local mass and stiffness of vocal folds. That is, both mass and stiffness increased along the superior-to-inferior direction. These variations were statistically analyzed under different experimental conditions (e.g., an increase in tension as a function of vocal fold elongation and an increase in stiffness and a decrease in mass as a function of glottal airflow). The study showed that physiologically relevant vocal fold tissue properties, which cannot be directly measured during in vivo human phonation, can be captured using this 3D-modeling technique.     

Material parameter computation for multi-layered vocal fold models

Today, the prevention and treatment of voice disorders is an ever-increasing health concern. Since many occupations rely on verbal communication, vocal health is necessary just to maintain one's livelihood. Commonly applied models to study vocal fold vibrations and air flow distributions are self sustained physical models of the larynx composed of artificial silicone vocal folds. Choosing appropriate mechanical parameters for these vocal fold models while considering simplifications due to manufacturing restrictions is difficult but crucial for achieving realistic behavior. In the present work, a combination of experimental and numerical approaches to compute material parameters for synthetic vocal fold models is presented. The material parameters are derived from deformation behaviors of excised human larynges. The resulting deformations are used as reference displacements for a tracking functional to be optimized. Material optimization was applied to three-dimensional vocal fold models based on isotropic and transverse-isotropic material laws, considering both a layered model with homogeneous material properties on each layer and an inhomogeneous model. The best results exhibited a transversal-isotropic inhomogeneous (i.e., not producible) model. For the homogeneous model (three layers), the transversal-isotropic material parameters were also computed for each layer yielding deformations similar to the measured human vocal fold deformations.     

利用声带动力学模型参数反演方法进行病变嗓音分类

提出一种声带动力学模型参数反演方法,从发声机理角度对声带病变嗓音进行有效区分。依据声带生理组织和伯努利定律构建声带动力学模型,确定模型优化参数向量,耦合声门气流获取模型声门波;利用迭代自适应逆滤波算法获得实际嗓音声门波作为目标声门波;采用遗传优化算法提出通过匹配目标和模型声门波特征参数实现模型参数反演。实验结果表明,表征声门波的各时频域参数匹配相对误差不超过2%;依据反演所获模型参数提出去除声门下压影响的平均归一化缩放系数,克服声带非对称性特征在区分病变嗓音方面的不足,实现病理嗓音的全面有效区分。      

Determination of superior surface strains and stresses, and vocal fold contact pressure in a synthetic larynx model using digital image correlation

Stresses and strains within the vocal fold tissue may play a critical role in voice fatigue, in tissue damage and resulting voice disorders, and in tissue healing. In this study, experiments were performed to determine mechanical fields on the superior surface of a self-oscillating physical model of the human vocal folds using a three-dimensional digital image correlation method. Digital images obtained using a high-speed camera together with a mirror system were used to measure displacement fields, from which strains, strain rates, and stresses on the superior surface of the model vocal folds were computed. The dependence of these variables on flow rate was established. A Hertzian impact model was used to estimate the contact pressure on the medial surface from superior surface strains. A tensile stress dominated state was observed on the superior surface, including during collision between the model folds. Collision between the model vocal folds limits the medial-lateral stress levels on the superior surface, in conjunction with compressive stress or contact pressure on the medial surface.     

Laryngeal Gestures During Stop Production Using High-Speed Digital Images

On acoustic and fiberscopic studies of stop consonants, voice onset time and glottal width have been shown to be greatest in heavily aspirated stops, next greater for slightly aspirated stops, and least for unaspirated stops. Integrated activity of the thyroarytenoid and posterior cricoarytenoid muscles has been reported to be involved in differentiating aspirate characteristics of the stops. However, the fine movement of mucosal edges of vocal folds during the production of stops has not been well documented. In recent years, a new method for high-speed digital recording of laryngeal dynamics has made this possible. In the current study, the movements of vocal fold edges were documented during the period of stop production using a fiberscopic system of high-speed digital images. By observing the glottal width and the visual vibratory movements of vocal folds before voice onset, the heavily aspirated stop was characterized as being more prominent and dynamic than the slightly aspirated and unaspirated stops.     

Voice source characteristics in Mongolian "throat singing" studied with high-speed imaging technique, acoustic spectra, and inverse filtering.

Mongolian "throat singing" can be performed in different modes. In Mongolia, the bass-type is called Kargyraa. The voice source in bass-type throat singing was studied in one male singer. The subject alternated between modal voice and the throat singing mode. Vocal fold vibrations were observed with high-speed photography, using a computerized recording system. The spectral characteristics of the sound signal were analyzed. Kymographic image data were compared to the sound signal and flow inverse filtering data from the same singer were obtained on a separate occasion. It was found that the vocal folds vibrated at the same frequency throughout both modes of singing. During throat singing the ventricular folds vibrated with complete but short closures at half the frequency of the true vocal folds, covering every second vocal fold closure. Kymographic data confirmed the findings. The spectrum contained added subharmonics compared to modal voice. In the inverse filtered signal the amplitude of every second airflow pulse was considerably lowered. The ventricular folds appeared to modulate the sound by reducing the glottal flow of every other vocal fold vibratory cycle.     

Voice Source Characteristics in Mongolian “Throat Singing” Studied with High-Speed Imaging Technique, Acoustic Spectra, and Inverse Filtering

Mongolian “throat singing” can be performed in different modes. In Mongolia, the bass-type is called Kargyraa. The voice source in bass-type throat singing was studied in one male singer. The subject alternated between modal voice and the throat singing mode. Vocal fold vibrations were observed with high-speed photography, using a computerized recording system. The spectral characteristics of the sound signal were analyzed. Kymographic image data were compared to the sound signal and flow inverse filtering data from the same singer were obtained on a separate occasion. It was found that the vocal folds vibrated at the same frequency throughout both modes of singing. During throat singing the ventricular folds vibrated with complete but short closures at half the frequency of the true vocal folds, covering every second vocal fold closure. Kymographic data confirmed the findings. The spectrum contained added subharmonics compared to modal voice. In the inverse filtered signal the amplitude of every second airflow pulse was considerably lowered. The ventricular folds appeared to modulate the sound by reducing the glottal flow of every other vocal fold vibratory cycle.     

A Novel Source-Filter Stochastic Model for Voice Production

The novel stochastic model to produce voiced sounds proposed in this paper uses the source-filter Fant theory to generate voice signals and, consequently, it does not consider the coupling between the vocal tract and the vocal folds. Two novelties are proposed in the paper. The first one is the new model obtained from the unification of two other deterministic one mass-spring-damper models obtained from the literature and the second one is to build a stochastic model which can generate and control the level of jitter resulting even in hoarse voice signals or with pathological characteristics but using a simpler model than those ones discussed in the literature. An inverse stochastic problem is then solved for two cases, considering a normal voice and other obtained from a case of paralysis on the vocal folds. The parameters of the model are identified in the two cases allowing the validation of the model.     

Chaos in voice, from modeling to measurement.

Chaos has been observed in turbulence, chemical reactions, nonlinear circuits, the solar system, biological populations, and seems to be an essential aspect of most physical systems. Chaos may also be central to the interpretation of irregularity in voice disorders. This presentation will summarize the results from a series of our recent studies. These studies have demonstrated the prescence of chaos in computer models of vocal folds, experiments with excised larynges, and human voices. Methods based on nonlinear dynamics can be used to quantify chaos and irregularity in vocal fold vibration. Studies have suggested that disordered voices from laryngeal pathologies such as laryngeal paralysis, vocal polyps, and vocal nodules might exhibit chaotic behaviors. Conventional parameters, such as jitter and shimmer, may be unreliable for analysis of periodic and chaotic voice signals. Nonlinear dynamic methods, however, have differentiated between normal and pathological phonations and can describe the aperiodic or chaotic voice. Chaos theory and nonlinear dynamics can enchance our understanding and therefore our assessment of pathological phonation.     

Restraining mechanisms in regulating glottal closure during phonation

Recent experimental studies showed that isotropic vocal fold models were often blown wide apart and thus not able to maintain adductory position, resulting in voice production with noticeable breathy quality. This study showed that the capability of the vocal fold to resist deformation against airflow and maintain adductory position can be improved by stiffening the body-layer stiffness or increasing the anterior-posterior tension of the vocal folds, which presumably can be achieved through the contraction of the thyroarytenoid (TA) and cricothyroid (CT) muscles, respectively. Experiments in both physical models and excised larynges showed that, when these restraining mechanisms were activated, the vocal folds were better able to maintain effective adduction, resulting in voice production with much clearer quality and reduced breathiness. In humans, one or more restraining mechanisms may be activated at different levels to accommodate the varying degree of restraining required under different voice conditions.     

Functional Analysis of Voice Using Simultaneous High-Speed Imaging and Acoustic Recordings

We present a comprehensive, functional analysis of clinical voice data derived from both high-speed digital imaging (HSDI) of the larynx and simultaneously acquired acoustic recordings. The goals of this study are to: (1) correlate dynamic characteristics of the vocal folds derived from direct laryngeal imaging with indirectly acquired acoustic measurements; (2) define the advantages of using a combined imaging/acoustic approach for the analysis of voice condition; and (3) identify new quantitative measures to evaluate the regularity of the vocal fold vibration and the complexity of the vocal output -- these measures will be key to successful diagnosis of vocal abnormalities. Image- and acoustic-based analyses are performed using an analytic phase plot approach previously introduced by our group (referred to as 'Nyquist' plot). Fast Fourier Transform (FFT) spectral analyses are performed on the same data for a comparison. Clinical HSDI and acoustic recordings from subjects having normal and specific voice pathologies, including muscular tension dysphonia (MTD) and recurrent respiratory papillomatosis (RRP) were analyzed using the Nyquist plot approach. The results of these analyses show that a combined imaging/acoustic analysis approach provides better characterization of the vibratory behavior of the vocal folds as it correlates with vocal output and pathology.     

A model for vocal fold vibratory motion, contact area, and the electroglottogram

The electroglottogram (EGG) has been conjectured to be related to the area of contact between the vocal folds. This hypothesis has been substantiated only partially via direct and indirect observations. In this paper, a simple model of vocal fold vibratory motion is used to estimate the vocal fold contact area as a function of time. This model employs a limited number of vocal fold vibratory features extracted from ultra high-speed laryngeal films. These characteristics include the opening and closing vocal fold angles and the lag (phase difference) between the upper and lower vocal fold margins. The electroglottogram is simulated using the contact area, and the EGG waveforms are compared to measured EGGs for normal male voices producing both modal and pulse register tones. The model also predicts EGG waveforms for vocal fold vibration associated with a nodule or polyp.     

Measures of spectral slope using an excised larynx model

Spectral measures of the glottal source were investigated using an excised canine larynx (CL) model for various aerodynamic and phonatory conditions. These measures included spectral harmonic difference H1-H2 and spectral slope that are highly correlated with voice quality but not reported in a systematic manner using an excised larynx model. It was hypothesized that the acoustic spectra of the glottal source were significantly influenced by the subglottal pressure, glottal adduction, and vocal fold elongation, as well as the resulting vibration pattern. CLs were prepared, mounted on the bench with and without false vocal folds, and made to oscillate with a flow of heated and humidified air. Major control parameters were subglottal pressure, adduction, and elongation. Electroglottograph, subglottal pressure, flow rate, and audio signals were analyzed using custom software. Results suggest that an increase in subglottal pressure and glottal adduction may change the energy balance between harmonics by increasing the spectral energy of the first few harmonics in an unpredictable manner. It is suggested that changes in the dynamics of vocal fold motion may be responsible for different spectral patterns. The finding that the spectral harmonics do not conform to previous findings was demonstrated through various cases. Results of this study may shed light on phonatory spectral control when the larynx is part of a complete vocal tract system.     

Symptomatology of adductor spasmodic dysphonia: A physiologic model

Vocal symptomatology of adductor spasmodic dysphonia (SD) is reviewed critically from historical, epidemiologic, and clinical perspectives. A model of symptomatology of this disorder based on a large patient population, and clinical and physiologic observations is advanced. The model incorporates crucial symptomatic and asymptomatic phonatory and nonphonatory physiologic parameters of laryngeal behavior in these patients. These parameters include vocal fold contact area, vocal fold collision force, glottic compression, and subglottic air pressure. Inappropriate efferent discharges from brain-stem basal ganglia are hypothesized as causing overadduction of the vocal folds in phonation, generating the basic and fundamental vocal symptom of adductor SD—strained, strangled, overpressured voice quality. Cortical loops are implicated as accountable for compensatory vocal behavior, not as the primary site of the disorder. Symptom occurrence, variability, magnitude, effects, and failure of treatment approaches, as well as recurrence of symptoms after ablative or invasive procedures, are explained by this model. The model also predicts that symptomatology of adductor spasmodic dysphonia is unique to this disorder and that symptoms are phonotopically organized. The minimal diagnostic battery based on the model is presented, and it is shown how this battery aids in the differential diagnosis of adductor SD and other phonatory disorders that closely mimic the vocal symptoms of adductor spasmodic dysphonia, including tremor.     

Effect of inferior surface angle on the self-oscillation of a computational vocal fold model

Geometry of the human vocal folds strongly influences their oscillatory motion. While the effect of intraglottal geometry on phonation has been widely investigated, the study of the geometry of the inferior surface of the vocal folds has been limited. In this study the way in which the inferior vocal fold surface angle affects vocal fold vibration was explored using a two-dimensional, self-oscillating finite element vocal fold model. The geometry was parameterized to create models with five different inferior surface angles. Four of the five models exhibited self-sustained oscillations. Comparisons of model motion showed increased vertical displacement and decreased glottal width amplitude with decreasing inferior surface angle. In addition, glottal width and air flow rate waveforms changed as the inferior surface angle was varied. Structural, rather than aerodynamic, effects are shown to be the cause of the changes in model response as the inferior surface angle was varied. Supporting data including glottal pressure distribution, average intraglottal pressure, energy transfer, and flow separation point locations are discussed, and suggestions for future research are given.     

Nonlinear dynamics of the voice: Signal analysis and biomechanical modeling

Irregularities in voiced speech are often observed as a consequence of vocal fold lesions, paralyses, and other pathological conditions. Many of these instabilities are related to the intrinsic nonlinearities in the vibrations of the vocal folds. In this paper, bifurcations in voice signals are analyzed using narrow-band spectrograms. We study sustained phonation of patients with laryngeal paralysis and data from an excised larynx experiment. These spectrograms are compared with computer simulations of an asymmetric 2-mass model of the vocal folds. (c) 1995 American Institute of Physics.