Quantitative Evaluation of Video Laryngostroboscopy: Reliability of the Basic Parameters

Transcription

1 Quantitative Evaluation of Video Laryngostroboscopy: Reliability of the Basic Parameters *Virgilijus Uloza, *Aurelija Vegien _e, *Ruta Pribuisien _e, and Viktoras Saferis, *ykaunas, Lithuania Summary: Objective. The purpose of this study was to evaluate quantitatively the basic parameters of the video laryngostroboscopy (VLS) and determine the sensitivity and specificity of these parameters discriminating healthy and pathological voice classes. Methods. Digital VLS recordings were performed for 159 individuals: 26 healthy and 133 patients. VLS variables (glottal closure, regularity, mucosal wave on the affected/healthy side, symmetry of vibration, and symmetry of image) were rated two times with the time interval of 1 year by three laryngologists. To evaluate interrater and test-retest reliability, intraclass correlation coefficients (ICCs) were calculated. To evaluate sensitivity and specificity of the VLS parameters, discriminant analysis was used. Results. Moderate to almost perfect levels (ICC ) of interrater reliability were revealed for most of the basic VLS parameters. The ICC of the interrater reliability was highest for symmetry of glottal image; the most problematic VLS parameter for rating was mucosal wave on the healthy side. ICC of the test-retest reliability were , P < An optimum system of VLS parameters discriminating normal and pathological voice subgroups with sensitivity 96.3% and specificity 100% included glottal closure and mucosal wave on the affected side. Conclusions. The quantitative evaluation of the VLS using basic parameters showed to be reliable in clinical settings and demonstrated high sensitivity and specificity distinguishing healthy and pathological voice patient groups. Key Words: Video laryngostroboscopy Quantification of VLS parameters Intra-class correlation coefficient. INTRODUCTION Video laryngostroboscopy (VLS) nowadays is considered as the most important and the most commonly used imaging method for vocal fold examination and evaluation of patients with voice disorders. 1 4 VLS evaluation allows early detection of infiltrative processes of the vocal folds, thus increasing potential possibilities of early diagnostics of laryngeal carcinoma. 5 8 This valuable imaging tool can also be used to assess the outcomes of therapy of laryngeal diseases or functional results of phonosurgical procedures However, the main limitation with the VLS remains the subjective nature of the interpretation of laryngeal phonatory function examination results, which significantly reduces the reproducibility and the use of VLS as a research tool or even as a quantitative instrument for comparing outcomes of treatment of voice and laryngeal disorders. 5,12 Therefore, current clinical decision making is typically based on an experienced, however, subjective evaluation of the VLS data. 13 Despite that the development of new digitalized and automated objective methods in VLS examination and/or videokymograhpy to enable precise determination of quantitative parameters is in progress, affordable techniques that allow to quantitatively assess the vibratory pattern of the vocal folds have not been available so far. 9,14 17 Accepted for publication December 17, From the *Department of Otolaryngology, Academy of Medicine, Lithuanian University of Health Sciences, Kaunas, Lithuania; and the ydepartment of Physics, Mathematics and Biophysics, Lithuanian University of Health Sciences, Kaunas, Lithuania. Address correspondence and reprint requests to Aurelija Vegien_e, Department of Otolaryngology, Academy of Medicine, Lithuanian University of Health Sciences, Eiveniu 2, LT , Kaunas, Lithuania. v_aurelija@yahoo.com Journal of Voice, Vol. 27, No. 3, pp /$36.00 Ó 2013 The Voice Foundation Several attempts have been made to elaborate various methods of measurement and quantification of VLS findings developing different VLS rating forms. 12,18 21 Various variables that have been evaluated during the VLS included fundamental frequency, symmetry of movements, periodicity, glottis closure, amplitude of vibration, mucosal wave, and the presence of nonvibrating portions of the vocal fold. 2,12,22 The number of VLS variables and peculiarities of the VLS parameters presented in literature differed from study to study; therefore, the data and results sometimes are hardly compatible. There is a lack of data in the literature about the specificity and sensitivity of VLS parameters discriminating normal and pathological voices. The aim of this study was to evaluate quantitatively the basic VLS parameters revealing their reliability and determine the sensitivity and specificity of these parameters discriminating healthy and pathological voice. MATERIAL AND METHODS The total group consisted of 159 individuals examined at the Department of Otolaryngology of Lithuanian University of Health Sciences, Kaunas, Lithuania, during the years A study group of 108 individuals consisting of a normal voice subgroup and a pathological voice subgroup underwent investigation for the reliability of the VLS parameters and sensitivity and specificity of VLS parameters separating normal voice and pathological voice groups. The normal voice subgroup comprised 26 randomly selected healthy volunteer individuals who considered their voice as normal. They had no complaints concerning their voice and no history of chronic laryngeal diseases or other long-lasting voice disorders. They also had never seen an otolaryngologist

2 362 Journal of Voice, Vol. 27, No. 3, 2013 for voice problems. The voices of this group of individuals were also evaluated as healthy voices by clinical voice specialists. No pathological alterations in the larynx of the subjects of the normal voice subgroup were found during VLS. Acoustic voice signal parameters of the normal voice subjects that were obtained using Voice Diagnostic Center lingwaves software, Version 2.5 (WEVOSYS, Forchheim, Germany) were within the normal range. The pathological voice subgroup consisted of 82 patients and represented a rather common, clinically discriminative group of laryngeal diseases, that is, mass lesions of vocal folds and paralysis. Mass lesions of vocal folds included in the study consisted of nodules, polyps, cysts, papillomata, keratosis, and carcinoma. Pathological voice group patients were recruited from the consecutive patients who were diagnosed with the laryngeal diseases mentioned previously. All the patients underwent clinical evaluation that included perceptual and acoustic voice assessments. The clinical diagnosis was based on typical clinical signs revealed during VLS and direct microlaryngoscopy. All the patients with mass lesions of vocal folds underwent endolaryngeal microsurgical interventions; therefore, the final diagnosis, which later served as gold standard, was proven by the results of histological examination of the removed tissue. VLS recordings of separate subgroup consisting of 90 patients with mass lesions of vocal folds and 34 glottal carcinoma patients were retrieved from the archive of the Department of Otolaryngology and this group was tested for sensitivity and specificity of VLS parameters discriminating carcinoma and mass lesions of vocal folds. Demographic data of the total study group and diagnoses of pathological voice group are presented in Table 1. Digital high quality VLS recordings were performed with an XION EndoSTROB DX device (XION GmbH, Berlin, Germany) using a 90 rigid endoscope. The subjects were seated for the VLS examination. No topical anesthesia of the oropharynx was used. The VLS examination and recordings were performed during modal phonation, that is, each subject was asked to sustain the vowel ee at a steady, comfortable pitch and loudness. Phonation time was kept long enough to allow for registration of a sustained phonation and at least one complete cycle of vibration. VLS recordings were rated independently by three experienced laryngologists/phoniatricians, using a standardized VLS assessment form consisting of six basic VLS parameters (Appendix). The following four standard VLS parameters were evaluated and quantified based on the protocol elaborated by the Committee on Phoniatrics of the European Laryngological Society: glottal closure, regularity of vibrations, mucosal wave, and symmetry of vibrations. Glottal closure including longitudinal, oval, hourglass-shaped gap; regularity of vibrations was defined as the degree to which one phonatory cycle suits the next; mucosal wave quantitative rating of the quality of the mucosal wave, accounting for the physiology of the layered structure of the vocal folds; symmetry quantitative rating of the mirror motion of both vocal folds. 22 Additionally, assessment of the mucosal wave was divided into: mucosal wave on the affected side and mucosal wave on the healthy side and the evaluation of symmetry was also divided in two parameters: symmetry of glottal image and symmetry of vibration. Thus, symmetry of glottal image includes symmetry of the anatomical structures and pathological lesions of the larynx. Symmetry of vibration reflects symmetry of glottal opening and closure during each cycle. For the rating of the VLS parameters, a 100-mm long visual analog scale (VAS) was used. A score of 0 (extreme left) meant normal perception of the parameter (no deviance), whereas 100 (extreme right) meant extreme deviance of the parameter evaluated. 22,23 Three experienced clinical voice specialists served as the raters. They performed perceptual blind evaluation and quantification of the VLS recordings without using any additional information about the subject s age, gender, diagnosis, and so forth. Before the evaluation, the experts were instructed about the goal of the study and agreed with the experimental quantification of the VLS parameters. Standardized instructions regarding viewing and evaluating the VLS recordings were given by one person. All raters agreed on sufficient quality of VLS recordings, which was acceptable for the rating. TABLE 1. Demographic Data and Diagnoses of Total Study Group (n ¼ 159) Gender Age (y) Diagnosis Number Female (n ¼ 77) Male (n ¼ 82) x ±SD Polyp ± 11.8 Nodules ± 11.5 Cyst ± 8.5 Reinke hyperplasia ± 8.5 Carcinoma T1 T ± 5.7 Vocal fold paralysis ± 13.8 Keratosis ± 14.2 Papillomatosis ± 21.4 Normal voice ± 10.9 Abbreviations: x, mean; SD, standard deviation.

3 Virgilijus Uloza, et al Quantitative Evaluation of VLS 363 To unify the VLS recordings rating, all 159 VLS recordings were presented to the experts in routine clinical manner, that is, with audio recording, and in a mixed and randomized order at one session. Each VLS sample was introduced to the raters and was followed by a short break to allow the experts to evaluate the sample. To approximate the conventional rating conditions, all recordings were presented in an ordinary room on a 21-inch screen. No contacts and discussions among the raters were allowed during the experiment. No repeat of examinations within the randomized session of review was allowed. Evaluation of the VLS recording on VAS for each patient took about 2 3 minutes. The same VLS recordings were evaluated two times with the time interval of 1 year. To verify the reliability of visual-perceptive measurements of distinct VLS parameter glottal closure objective analysis of frozen VLS images was performed. Relevant parts of VLS recordings were stored with a frame grabber Any Video Converter.lnk (AVCLabs, Inc., Shenzhen, China). Contours of glottal area, both during maximum opening and maximum closure were traced manually with a mouse and filled in with color (Figure 1). Thus, the colored areas of the glottis were computed in pixels using the Laringometrija (Elinta Inc., Kaunas, Lithuania) program. 24 A correct comparison of images of separate VLS recordings of different larynges requires equal magnifications and view angles in all images. However, in clinical settings, the distances between the optical lens of the rigid endoscope and the level of vocal folds are rather different. Therefore, to minimize the influence of these differences, the relative measurement, that is, relative glottal area (RGA) was calculated (RGA ¼ GAmax/ GAmin; where GAmax, glottal area of the maximum opening and GAmin, glottal area of maximum closure). Spearman correlation analysis was used to proof concordance of RGA, GAmin data, and VAS measurements of VLS glottal closure. STATISTICS Statistical analysis was performed using IBM SPSS Statistics for Windows, version 20.0 (Armonk, NY: IBM Corp.). Data were presented as mean (x) ± standard deviation (SD). The Student t test was used for testing hypotheses about equality of means. For testing hypotheses about the independence, the chi-square test was used. The intraclass correlation coefficient (ICC) was used for observer agreement assessment. The ICC was defined as the ratio of the variance between subjects to the total variance. 25,26 These variances are derived from analyses of variance (ANOVAs). The ICC (from two-way models) was used for the assessment of interrater agreement, and ICC (from one-way random effects ANOVA model) was used for the assessment of intrarater (test-retest) consistency. ICC values were interpreted as follows: <0.2, poor; , fair; , moderate; , substantial (strong); and >0.80, almost perfect agreements. The correlation between VLS parameters and objective measurements of glottal area was tested using Spearman correlation coefficient (rho). Discriminant analysis was used to determine limiting values of VLS parameters discriminating normal and pathological voice groups. Sensitivity and specificity of VLS parameters were revealed from classification tables. The level of statistical significance by testing statistical hypothesis was RESULTS Interrater reliability of VLS parameters Generally, the ICC obtained from the total study group represented statistically significant (P ¼ 0.000) substantial to almost perfect interrater reliability for five VLS parameters evaluated. Some exception has been revealed only for the mucosal wave on the healthy side (Table 2). Table 2 shows the ICC for interrater reliability. The ICC obtained during the first session and during the second session 1 year after remained on the same level and did not differ significantly, thus confirming substantial to almost perfect interrater reliability. The interrater ICC for the mucosal wave on the healthy side improved from 0.34 fair during the first session to 0.69 substantial during the second session. Table 3 presents the ICC reflecting the interrater reliability for the pathological voice subgroup. All VLS parameters evaluated in the study revealed statistically significant ICC for interrater reliability, both during the first session and during the second session 1 year after. The lowest, however, statistically significant ICC was found for the mucosal wave on the healthy side demonstrating poor reliability of the parameter. However, during the second session, the ICC for this FIGURE 1. VLS image: carcinoma of the left vocal fold. (A) Respiration (maximum glottal opening): white-colored area, GAmax. (B) Phonation (maximum glottal closure): white-colored area, GAmin.

4 364 Journal of Voice, Vol. 27, No. 3, 2013 TABLE 2. ICC for the Interrater Reliability During the First and the Second Sessions (Normal and Pathological Voice Subgroups; n ¼ 108) VLS Parameters ICC-1 Interrater P ICC-2 Interrater P Glottal closure Regularity Mucosal wave on the affected side Mucosal wave on the healthy side Symmetry of vibration Symmetry of glottal image Abbreviations: ICC, intra-class correlation coefficient; ICC-1, the first session; ICC-2, the second session 1 year after. parameter reached a moderate level of reliability. All ICC remained statistically significant during the second session 1 year after. Intrarater (test-retest) reliability of VLS parameters Table 4 presents the ICC reflecting the intrarater reliability as the result of test-retest investigation. The ICC calculated for individual raters demonstrated substantial to almost perfect intrarater (test-retest) reliability for all raters. Some exception presented only mucosal wave on the healthy side (ICC ¼ 0.34) for the first rater. Sensitivity and specificity of VLS parameters The mean values and SDs of the VLS parameters in normal voice and pathological voice groups are presented in Table 5. A statistically significant difference (P < 0.001) between patients group and the normal voice group was found of all VLS parameters measured, with the patients having worse results, thus demonstrating measurable evidence of the differences between a normal and a deteriorated phonation pattern. Table 6 presents data of sensitivity, specificity, and limiting values of the VLS parameters discriminating normal and pathological voice subgroups. The highest sensitivity and specificity separating normal and pathological voice subgroups was revealed for the mucosal wave on the affected side (sensitivity 95.1%, specificity 100%, and limiting value 42.6 VAS points). Other VLS parameters presented with high sensitivity in the limits of %, whereas the specificity reached 100% for all VLS parameters measured. The 100% level of the specificity of the VLS parameters discriminating normal and pathological voice subgroups could be determined by clinically evident deterioration of phonation pattern in all cases of mass lesions of vocal folds or paralysis and as a consequence deviation of the VLS parameters. The limiting values of the VLS parameters were determined as the optimum point for separating normal and pathological voice subgroups. Therefore, parametric values of the VLS parameters that were found to be larger than the limiting values were considered as pathological. As the result of discriminant analysis using mean values of the VLS parameters, an optimum system of VLS parameters discriminating normal and pathological voice subgroups was established. This system included two VLS variables: glottal closure and the mucosal wave on the affected side. The sensitivity of this system was 96.3% and specificity 100%. Results of sensitivity and specificity of the VLS parameters discriminating laryngeal carcinoma T1 T2 and other mass lesions of vocal folds are presented in Table 7. The highest sensitivity (91.2%) separating glottal carcinoma T1 T2 and other mass lesions of vocal folds were revealed for the mucosal wave on the affected side, however, the specificity was only 51.1% and the limiting value was 84.3 VAS points. Other VLS parameters showed sensitivity in the limits of %, whereas the specificity was in the limits of %. As the result of discriminant analysis using mean values of the VLS parameters, an optimum system of VLS parameters discriminating carcinoma T1 T2 and other mass lesions of vocal folds groups was established. This system included the four following VLS variables: Regularity, mucosal wave on the affected side, symmetry of vibration, and symmetry of glottal image. The sensitivity of this system was 85.3% and specificity 74.4%. TABLE 3. ICC for the Interrater Reliability During the First and the Second Sessions (Pathological Voice Subgroup; n ¼ 82) VLS Parameters ICC-1 Interrater P ICC-2 Interrater P Glottal closure Regularity Mucosal wave on the affected side Mucosal wave on the healthy side Symmetry of vibration Symmetry of glottal image Abbreviations: ICC, intra-class correlation coefficient; ICC-1, the first session; ICC-2, the second session 1 y after.

5 Virgilijus Uloza, et al Quantitative Evaluation of VLS 365 TABLE 4. ICC for the Intrarater Reliability (Test-Retest) VLS Parameters ICC (Test-Retest) Rater 1 P Rater 2 P Rater 3 P Glottal closure Regularity Mucosal wave on the affected side Mucosal wave on the healthy side Symmetry of vibration Symmetry of glottal image P, statistically significant test-retest ICC for each separate rater. Abbreviation: ICC, intra-class correlation coefficient. Correlation analysis revealed statistically significant (P < 0.001) moderate correlations among the VLS parameter glottal closure and objectively measured RGA (r ¼ 0.54) and GAmin (r ¼ 0.51), thus confirming the reliability and sufficient correctness of that visual-perceptual VLS measurement. DISCUSSION During decades, VLS remains the ordinary and inalterable method in clinical practice and research as the only clinically feasible test that allows visualization of the vocal folds and their vibratory function. The demand for objective and quantitative image analysis in clinical practice and research is obvious. Quantitative measures of the VLS variables could provide more reliable information for diagnostic requirements and be useful in monitoring a patient s treatment progress over time. 13 However, subjective ratings of amplitude, periodicity, regularity of vocal fold vibration, mucosal wave, and other VLS parameters face rather essential shortcomings because of the subjectivity of evaluation and as a consequence of differences in assessment. Despite these deprivations of VLS assessment, some previous studies have shown that several VLS measures of the vibratory pattern of the vocal folds are clinically relevant and quantifiable. 9,12,19 On the other hand, Shneider et al 27 revealed that quantitative measurements of selected VLS parameters (three geometrical and three time dependent) did not correlate with qualitative subjective stroboscopic assessment in functional dysphonias. Because assessment of the VLS is based on the visualperceptual judgments, achievement of acceptable levels of interrater and intrarater reliability is of critical importance in clinical settings and research. Data presented in the literature on this matter showed rather wide limits of reliability ranging from 0.2 to ,21,28 However, substantial to almost perfect levels of interrater and intrarater reliability of the most measurements of VLS variables revealed in the present study confirmed feasibility of VAS based on the quantification of the basic VLS parameters. Phonation is an aerodynamic and acoustic phenomenon of an extremely complex vibratory system, which is not a perfect machine. Part of the evidence for this complexity is the fact that cycle-to-cycle consistency of vocal folds vibrations in phonation is never completely regular and even normal human voices deviate slightly from perfect periodicity, as well as the pattern of glottal closure may slightly differ from a perfect closure. 29,30 This natural imperfectness of phonation pattern was partly reflected in the present study by the mean values of the VLS parameters evaluated on VAS, which demonstrated slight deviations from possible ideal normal perception of the parameters even in cases of a normal voice (Table 5). Pathological vocal fold changes including mass lesions may interfere with the glottal vibratory pattern causing perturbations in TABLE 5. Mean Values of VLS Parameters in Normal and Pathological Voice Subgroups VLS Parameters VAS Points Normal Voice Subgroup, n ¼ 26 Pathological Voice Subgroup, n ¼ 82 x ±SD Glottal closure 8.4 ± ± Regularity 6.9 ± ± Mucosal wave on the affected side 4.2 ± ± Mucosal wave on the healthy side 4.2 ± ± Symmetry of vibration 1.3 ± ± Symmetry of glottal image 1.4 ± ± Abbreviations: x, mean; SD, standard deviation. x ±SD P

6 366 Journal of Voice, Vol. 27, No. 3, 2013 TABLE 6. Sensitivity and Specificity of the VLS Parameters Discriminating Normal and Pathological Voice Groups (n ¼ 108) VLS Parameters Sensitivity, % Specificity, % Limiting Value (VAS Points) Glottal closure Regularity Mucosal wave on the affected side Mucosal wave on the healthy side Symmetry of vibration Symmetry of glottal image regularity. As a consequence, deviations of the periodicity are often greater in pathological than in healthy voices. 31,32 Disruption of the laminar structure of the vocal fold by neoplastic tissue changes the mass, stiffness, geometry of the affected vocal fold, and symmetry of glottis resulting in a reduced mucosal wave. 5,7 In the present study, mean VLS parameters regularity, mucosal wave, symmetry of vibration, and symmetry of glottal image reflecting both pathological patterns mentioned previously were found significantly higher in the pathological voice subgroup compared with the subgroup of the normal voice confirming theoretical considerations mentioned previously. Prevention of full glottal approximation by the intrusion of an additional surface mass can even more deteriorate the phonation pattern. Again, this was decisively demonstrated in this study by statistically significantly increased mean values of incomplete glottal closure in the pathological voice group (sensitivity 91.5% and specificity 100%). Moreover, both accuracy and reliability of that VLS parameter were confirmed in this study by objective measurement of RGA and GAmin. In this study, moderate-to-high rates of sensitivity ( %) and moderate rates of specificity ( %) of the basic VLS parameters discriminating laryngeal carcinoma and other mass lesions of vocal folds were revealed. Some exception presented mucosal wave on the affected side, which showed very high 91.2% sensitivity and rather low 51.1% specificity. However, optimum system of four following VLS variables regularity, mucosal wave on the affected side, symmetry of vibration, and symmetry of glottal image increased the sensitivity up to 85.3% and specificity up to 74.4%. However, this demonstrates that even after aggregation of several of VLS parameters, the VLS VAS rating alone does not reach reliable level for clinical utility discriminating laryngeal carcinoma. Generally, this is in concordance with the data of the literature, as the former studies showed that the pathological VLS features associated with keratosis, that is, reduced or abolished mucosal wave propagation that has been considered as a possible predictor for the presence of carcinoma, do not reliably predict the presence of carcinoma or depth of invasion by the tumor. 5,7 Therefore, a combination of VLS with other diagnostic procedures is advocated. Peretti et al 6 in their study showed 82% specificity and 100% sensitivity of VLS in diagnostic of glottal carcinoma; however, the combination of VLS with saline infusion into Reinke space raised the values of specificity up to 89% and sensitivity up to 100%. Gugatschka et al 8 demonstrated that the combination of cytology and pathological VLS allowed detection of glottic cancer with a sensitivity of 97%, in contrast to 74% as found by cytology alone. Another way to increase credibility of the VLS examination could be elaboration of automated systems of images analysis including additional types of features to characterize color, texture, and geometry of biological structures seen in color laryngeal images. 33 Simultaneously, development of visualperceptual quantitative rating systems of various VLS variables would allow quantification of VLS; thus it could serve as a clinically feasible instrument and enlarge the probability for the development of automated VLS analysis systems. The numerical values of the VLS parameters obtained in this study have not contributed much to the pathological categorization of laryngeal diseases; however, they depended on the severity of laryngeal pathology and voice disorder. Additional new TABLE 7. Sensitivity and Specificity of the VLS Parameters Discriminating Laryngeal Carcinoma and Mass Lesions of Vocal Folds Subgroups VLS Parameters Glottal Carcinoma T1 T2 (n ¼ 34) and Mass Lesions of Vocal Folds (n ¼ 90) Sensitivity, % Specificity, % Limiting Value (VAS Points) Glottal closure Regularity Mucosal wave on the affected side Mucosal wave on the healthy side Symmetry of vibration Symmetry of glottal image

7 Virgilijus Uloza, et al Quantitative Evaluation of VLS 367 VLS variables included in this study, mucosal wave on the healthy/affected side and symmetry of vibration/glottal image, demonstrated substantial reliability. Therefore, the use of these new VLS variables the additional subset evaluation of mucosal wave (with separate evaluation of the healthy and affected side) and symmetry (separate evaluation of glottal image and vibration) in clinical practice would increase the completeness of the VLS examination. The limitations of quantification of VLS variables originating from inherent restrictions of stroboscopic examination must be considered. The stroboscopic image of the vocal fold vibration is basically an optical illusion made by the human eye, arising from the virtual reconstruction of adjacent phases of periodic oscillatory samples. Therefore, VLS recordings of aphonic patients and/or nonperiodic voices may be difficult to analyze identifying and quantifying variables of the vocal folds vibratory pattern. Moreover, some patients with large obstructive laryngeal lesions and difficult-to-visualize larynges or poorly compliant individuals will not be suitable for classical VLS recordings and assessment. 9 As a consequence, some of the subtle and meticulous classical VLS variables as phase closure, phase symmetry, and regularity could not be determined and assessed. Recent study of Nawka and Konerding 28 showed that, for example, the interrater reliabilities for vertical level, glottal closure, phase closure, phase symmetry, and regularity are so low that these variables should not be assessed via stroboscopy. On the other hand, the basic VLS parameters analyzed in this study are rather simply recognized, they also involve nonperiodic findings (lack or absence of mucosal wave); therefore, they were found to be clinically relevant and reliable. Another limitation arises from the study design. The study group consisted of clinically obvious vocal fold lesions, which demonstrated high sensitivity and specificity of VLS VAS rating discriminating normal and pathological voice subgroups and this result could be anticipated. Therefore, the results of the present study should be considered as a preliminary evaluation of the VAS scoring of the stroboscopy parameters. Further research to prove clinical relevance of VLS rating including primary muscular tension dysphonia and assessment of phonosurgical and/or therapeutical treatment of voice disorders is required. Investigation of possible correlations of VLS VAS scores to acoustic voice measures and vocal quality scales (Voice Handicap Index and Grade, Roughness, Breathiness, Asthenia, Strain) would be of great value importance to prove clinical validity of VAS-based VLS rating. Finally, one can assume that quantification with VAS is not a true quantification; it is a qualitative measure of subjective perception. The real quantification of VLS could be attained in future development of automatic VLS analysis systems. Nevertheless, results of this study could be helpful in sampling of VLS parameters for the development of such systems. CONCLUSIONS In this study, we have demonstrated that (1) quantification of basic VLS parameters is possible, rather simple, reliable, and clinically feasible; (2) numerical values of VLS parameters of pathological voices demonstrate significant deviances from these in normal voices; (3) quantification of basic VLS parameters is rather sensitive and specific discriminating normal and pathological voices. The VLS rating form presented in this study is relatively simple, fast, does not require very special training and therefore is feasible for clinical use providing documentation and reproducibility of the VLS examination. This method can also be used for research in which the rating of the basic VLS parameters is required. Quantification of the VLS parameters widens the possibilities for further investigation on correlations of VLS variables and acoustic voice parameters. Summarizing, the quantitative evaluation of the VLS using basic parameters showed to be reliable in clinical settings and demonstrated high sensitivity and specificity distinguishing healthy and pathological voice patients groups. REFERENCES 1. Hirano M, Bless DM. Videostroboscopic Examination of the Larynx. San Diego, CA: Singular Publishing Group; 1993: Sataloff RT. Professional Voice: The Science and Art of Clinical Care. 3rd ed. San Diego, CA: Plural Publishing, Inc.; 2005: Pribuisiene R, Uloza V, Kupcinskas L, Jonaitis L. Perceptual and acoustic characteristics of voice changes in reflux laryngitis patients. J Voice. 2006; 20: Yelken K, Gultekin E, Guven M, Eyibilen A, Aladag I. Impairment of voice quality in paradoxical vocal fold motion dysfunction. J Voice. 2010;24: Colden D, Zeitels S, Hillman R, Jarboe J, Bunting G, Spanou N. Stroboscopic assessment of vocal-fold atypia and early cancer. Ann Otol Rhinol Laryngol. 2001;110: Peretti G, Piazza C, Berlucchi M, Cappiello J, Giudice M, Nicolai P. Preand intraoperative assessment of mid-cord erythroleukoplakias: a prospective study on 52 patients. Eur Arch Otorhinolaryngol. 2003;260: Gamboa J, Echeverrıa L, Molina B, Cobeta I. Stroboscopic assessment of chronic laryngitis. Acta Otorrinolaringol Esp. 2006;57: Gugatschka M, Kiesler K, Beham A, Rechenmacher J, Friedrich G. Hyperplastic epithelial lesions of the vocal folds: combined use of exfoliative cytology and laryngostroboscopy in differential diagnosis. Eur Arch Otorhinolaryngol. 2008;265: Noordzij JP, Woo P. Glottal area waveform analysis of benign vocal fold lesions before and after surgery. Ann Otol Rhinol Laryngol. 2000;109: Uloza V, Pribuisiene R, Saferis V. Multidimensional assessment of functional outcomes of medialization thyroplasty. Eur Arch Otorhinolaryngol. 2005;262: Mesallam TA, Khalil YA, Malki KH, Farahat M. Medialization thyroplasty using autologous nasal septal cartilage for treating unilateral vocal fold paralysis. Clin Exp Otorhinolaryngol. 2011;4: Rosen CA. Stroboscopy as a research instrument: development of a perceptual evaluation tool. Laryngoscope. 2005;115: Dailey SH, Kobler JB, Hillman RE, Tangrom K, Thananart E, Mauri M, Zeitels SM. Endoscopic measurement of vocal fold movement during adduction and abduction. Laryngoscope. 2005;115: Schuberth S, Hoppe U, Dollinger M, Lohscheller J, Eysholdt U. Highprecision measurement of the vocal fold length and vibratory amplitudes. Laryngoscope. 2002;112(6): Speyer R, Wieneke GH, Kersing W, Dejonckere PH. Accuracy of measurements on digital videostroboscopic images of the vocal folds. Ann Otol Rhinol Laryngol. 2005;114: Deguchi S, Ishimaru Y, Washio S. Preliminary evaluation of stroboscopy system using multiple light sources for observation of pathological vocal fold oscillatory pattern. Ann Otol Rhinol Laryngol. 2007; 116:

8 368 Journal of Voice, Vol. 27, No. 3, Manfredi C, Bocchia L, Cantarella G, Peretti G. Videokymographic image processing: objective parameters and user-friendly interface. Biomed Signal Process Contr. 2012;7: Sercarz JA, Berke GS, Arnstein D, Gerratt B, Natividad M. A new technique for quantitative measurement of laryngeal videostroboscopic images. Arch Otolaryngol Head Neck Surg. 1991;117: Woo P. Stroboscopy. San Diego, CA: Plural Publishing, Inc.; 2010: Dejonckere PH, Crevier L, Elbaz E, Marraco M, Millet B, Remacle M, Woisard V. Quantitative rating of video-laryngostroboscopy: a reliability study. Rev Laryngol Otol Rhinol (Bord). 1998;119: Poburka BJ. A new stroboscopy rating form. J Voice. 1999;13: Dejonckere PH, Bradley P, Clemente P, et al. A basic protocol for functional assessment of voice pathology, especially for investigating the efficacy of (phonosurgical) treatments and evaluating new assessment techniques. Guideline elaborated by the Committee on Phoniatrics of the European Laryngological Society (ELS). Eur Arch Otorhinolaryngol. 2001;258: Shiffman S, Reinolds MI, Young FW. Introduction to Multidimensional Scaling: Theory, Methods and Applications. New York: Academic Press; 1981: Kaseta M. Development and Assessment of Automated Categorization System for Organic Laryngeal Diseases [dissertation]. [Kaunas, Lithuania]: Kaunas University of Medicine; 2010: Shrout PE, Fleiss JL. Intraclass correlation: uses in assessing rater reliability. Psychol Bull. 1979;88: de Vet H. Observer reliability and agreement. In: Armitage P, Colton T, eds. Encyclopedia of Biostatistics. 2nd ed. Chichester, UK: Wiley; 2010: Schneider B, Wendler J, Seidner W. The relevance of stroboscopy in functional dysphonias (1). Folia Phoniatr Logop. 2002;54: Nawka T, Konerding U. The interrater reliability of stroboscopy evaluations. J Voice. 2012;26:812.e1 812.e10 [Epub ahead of print]. 29. Sulter AM, Schutte HK, Miller DG. Standardized laryngeal videostroboscopic rating: differences between untrained and trained male and female subjects, and effects of varying sound intensity, fundamental frequency, and age. J Voice. 1996;10: Brockmann M, Drinnan MJ, Storck C, Carding PN. Reliable jitter and shimmer measurements in voice clinics: the relevance of vowel, gender, vocal intensity, and fundamental frequency effects in a typical clinical task. J Voice. 2011;25: Vieira MN, McInnes FR, Jack MA. On the influence of laryngeal pathologies on acoustic and electrographic jitter measures. J Acoust Soc Am. 2002; 111: Petrovic-Lazic M, Babac S, Vukovic M, Kosanovic R, Ivankovic Z. Acoustic voice analysis of patients with vocal fold polyp. J Voice. 2011;25: Verikas A, Gelzinis A, Bacauskiene M, Hallanderb M, Uloza V, Kaseta M. Combining image, voice, and the patient s questionnaire data to categorize laryngeal disorders. Artif Intell Med. 2010;49: APPENDIX