- Email: [email protected]
Vocal Fold Vibration Following Surgical Intervention in Three Vocal Pathologies: A Preliminary Study
ARTICLE IN PRESS Vocal Fold Vibration Following Surgical Intervention in Three Vocal Pathologies: A Preliminary Study *Wenli Chen, †Peak Woo, and ‡Thomas Murry, *Phoenix, Arizona, †New York, New York, and ‡Loma Linda, California Summary: High-speed videoendoscopy captures the cycle-to-cycle vibratory motion of each individual vocal fold in normal and severely disordered phonation. Therefore, it provides a direct method to examine the specific vibratory changes following vocal fold surgery. The purpose of this study was to examine the vocal fold vibratory pattern changes in the surgically treated pathologic vocal fold and the contralateral vocal fold in three vocal pathologies: vocal polyp (n = 3), paresis or paralysis (n = 3), and scar (n = 3). Digital kymography was used to extract high-speed kymographic vocal fold images at the mid-membranous region of the vocal fold. Spectral analysis was subsequently applied to the digital kymography to quantify the cycle-to-cycle movements of each vocal fold, expressed as a spectrum. Surgical modification resulted in significantly improved spectral power of the treated pathologic vocal fold. Furthermore, the contralateral vocal fold also presented with improved spectral power irrespective of vocal pathology. In comparison with normal vocal fold spectrum, postsurgical vocal fold vibrations continued to demonstrate decreased vibratory amplitude in both vocal folds. Key Words: High-speed videoendoscopy–Digital kymography–Vocal fold vibration–Laryngeal surgery–Vocal fold vibratory spectrum.
INTRODUCTION Postoperative assessment of vocal fold vibration is imperative from a surgical perspective as well as a rehabilitative perspective. It determines surgical success, candidacy for voice therapy, and whether further medical management is needed. Current postoperative assessments of vocal fold vibration consist of subjective analysis utilizing videostroboscopy, self-assessment tools, or audio or aerodynamic instrumental analysis. In many cases, the patient is left to making a simple self-perception judgment of “better, worse, or the same.” High-speed videoendoscopy (HSV) captures data at more than 2000 frames per second, which allows direct visualization of the entire cycle-to-cycle vibratory motion of the left and right vocal fold in normal speaking adults1–8 and in disordered phonation.9–15 The use of HSV has provided significant utility in understanding the etiology of disorders on vocal fold vibration, as well as provided close examination of aperiodic vibrations that are difficult to capture on standard stroboscopy measures.9–15 Digital kymography uses HSV to examine the precise vibratory characteristics of each vocal fold at selected locations along the vocal folds.7,9,14,15 The resulting values can be analyzed as a vibratory spectrum.7,9 In the vibratory spectrum, spectral power in the fundamental frequency (F0 = H1) has been associated with the degree of vocal fold excursion, whereas the energy of the higher harmonics has been associated with the discontinuity that occurs with vocal fold impact.9,16 Therefore, digital kymography spectrum is a useful tool to objectively quantify precise vocal
Accepted for publication February 14, 2017. A version of the paper was presented at the 2015 Fall Voice Conference in Pittsburgh, PA on October 17, 2015. From the *Department of Otorhinolaryngology, Mayo Clinic, Phoenix, Arizona; †Department of Otolaryngology Head and Neck Surgery, Icahn School of Medicine, New York, New York; and the ‡Department of Otolaryngology Head and Neck Surgery, Loma Linda University, Loma Linda, California. Address correspondence and reprint requests to Wenli Chen, Mayo Clinic, 5779 E Mayo Blvd, Phoenix, AZ 85054. E-mail: [email protected] Journal of Voice, Vol. ■■, No. ■■, pp. ■■-■■ 0892-1997 © 2017 The Voice Foundation. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jvoice.2017.02.006
fold vibration and offers an ideal methodology to examine the specific changes in vibratory behavior of the vocal folds following surgical intervention.9 Quantification of vibration in each vocal fold using HSV is now practical. Previous studies using HSV demonstrated improvements in vocal fold vibration following medical intervention. Kunduk et al11 have shown improved amplitude and symmetry of vocal fold vibration following surgical resection of vocal fold polyp at 1 month and 3 months postsurgery. Kimura et al13 demonstrated improvements in vocal vibratory patterns toward symmetry in patients with unilateral vocal fold paralysis following collagen injection. Chen et al9 have shown improved vibratory amplitude and symmetry in three types of vocal fold pathologies following surgical treatment, with the most remarkable change noted at the mid-membranous region of the vocal fold. Although these studies unequivocally demonstrate improved vibratory motion in both vocal folds, little is known about the specific vibratory changes in each vocal fold. Understanding changes in the treated as well as the contralateral unoperated vocal fold may assist in surgical management and offer insight as to the basis of remaining dysphonia or the lack of it. Therefore, the goal of this preliminary study was to examine direct vocal fold vibratory changes in the treated pathologic vocal fold and the contralateral vocal fold following surgical intervention. METHODS Three subjects with identifiable mid-vocal fold benign lesions, two with unilateral vocal fold paresis, one with unilateral vocal fold paralysis, and three with vocal fold scar were studied prior to and following surgical intervention. Vocal pathology was diagnosed by a board-certified otolaryngologist. These diagnostic types were intentionally selected to represent varying levels of vocal fold pathology (refer to Table 1 for laryngeal etiology, age, diagnosis, and the type of surgical intervention). Two male and one female normal adult subjects were also recorded for comparison. HSV was obtained using standard videostroboscopy procedures.
ARTICLE IN PRESS 2
Journal of Voice, Vol. ■■, No. ■■, 2017
TABLE 1. Information on Pathologic Group, Age, Diagnosis, and Type of Surgical Intervention Received Subject M1 M2 M3 S1 S2 S3 P1 P2 P3
Age
Diagnosis
Diagnostic Information
Surgical Intervention
42 45 47 85 71 68 45 49 67
Polyp Polyp Polyp Scar Scar Scar Paralysis Paresis Paresis
L VF polyp R VF polyp L VF polyp R VF hyperkeratosis L VF scar with recurrent keratosis L VF scar with R VF bowing R VF paralysis with R sulcus vocalis L VF paresis L VF paresis
Polyp excision using cold instruments Polyp excision using cold instruments KTP laser KTP laser KTP laser KTP laser Injection laryngoplasty Injection laryngoplasty Injection laryngoplasty
Abbreviation: KTP, potassium titanyl phosphate; L, left; R, right; VF, vocal fold.
High-speed video recording HSV was recorded with the Kay Elemetrics High-Speed Digital Imaging (HSDI) system (KayPENTAX Photron Motion, Montvale, NJ, USA), which consisted of a 90° rigid endoscope (Model 9100) and a 300-Watt Xenon light source. The HSDI system captured grayscale images at a rate of 2000 frames per second, with a spatial resolution of 256 × 120 pixels rotated to a vertical position. Videostrobolaryngoscopy was conducted utilizing a rigid endoscope, with the subject sustaining /i/ at a comfortable pitch and loudness. A contact microphone was attached at the neck to monitor pitch and a second microphone was placed 6 inches from the lips to monitor intensity. The HSV samples were obtained when the examiner observed a clear and full view of the larynx. Six continuous 2-second tokens of the vowel were recorded. The three best tokens with a clear and full view of the vocal folds were saved onto a hard drive for analysis. All subjects tolerated the data collection procedure without any difficulties. Data analysis Kymography image processing HSV images were preprocessed with video editing software (VirtualDub v.1.9.11 (virtualdub.org)). Brightness and contrast of the glottis and vocal folds were adjusted for optimal edge detection. Image rotation was implemented, as needed, to ensure vertical alignment of the image. A 400- to 500-frame video segment that captured a full view of the vocal folds with minimal movement of the subject was extracted from the recorded HSV samples. Kay’s
Image Processing Software (KIPS, Model 9181) was used to generate the kymogram. Digital kymograph was created by placing a transverse line across the mid-membranous region of the glottis (Figure 1A and B), which has been reported to be the area of maximal vocal fold contact.17 Edge detection was applied to trace the vocal fold edges (Figure 1C). Kymograph analysis of the vibratory samples is dependent on the delineation of the vocal fold edge from HSV.18 Therefore, manual corrections function was utilized, as needed, to ensure correct tracing of the vocal fold edges. Kymograph edge analysis function was subsequently applied on the kymogram. The resulting values were kymograph edge data, which showed the coordinate values of the left and right edges of the vocal fold across time (Figure 1D). Fourier transform function was subsequently applied. This resulted in a spectrum ranging from 0 Hz to 1000 Hz for the left and right edges of the vocal folds (Figure 1E). Spectral data analyses Three values were extracted from the left and the right vocal fold vibratory spectrum for quantitative analysis: the peak power values of the fundamental (F0 = H1), second harmonic (H2), and third harmonic (H3). Twenty-five percent of the postprocessed 500frame HSV tokens were randomly selected and reanalyzed by the same experimenter to evaluate the error of measurement. Comparison of the H1, H2, and H3 values between the original and reanalyzed sample yielded a reliability of 87%. The following vocal fold vibratory parameters were extracted at baseline and postintervention: (1) total spectral power,
FIGURE 1. Methods to obtain spectral analysis of digital kymography in subject P2 (see text for descriptions).
ARTICLE IN PRESS Wenli Chen, et al
Vocal Fold Vibration Following Surgical Intervention
3
FIGURE 2. Mean values of total spectral power prior to and following surgical intervention in the treated pathologic vocal fold (Tx VF) and the untreated contralateral vocal fold (contra VF).
FIGURE 3. Mean spectral values of F0, H2, and H3 prior to and following surgical intervention in the treated pathologic vocal fold (treated VF) and the untreated contralateral vocal fold (contra VF).
(2) spectrum shape, and (3) the Voice Handicap Index-10 (VHI10). The total spectral power was determined by calculating the sum of the spectral peak values of H1, H2, and H3 altogether. Spectrum shape was determined by tabulating the spectral power at H1, H2, and H3 individually. Lastly, VHI-10 was administered to examine subjects’ perceptions of their voice handicap. Statistical analysis was performed with t test using the JMP software (JMP Pro 10, SAS Institute, Cary, NC, USA).
with a unilateral vocal fold paresis or paralysis, a marked increase in spectral power was noted in the treated vocal fold as compared with the contralateral vocal fold. Subjects with a vocal fold scar presented with a marked increase in the spectral power of the contralateral vocal fold as compared with the treated vocal fold. The greatest change in the total spectral power was noted in the polyp group. The least change was noted in the scar group.
RESULTS Total spectral power Prior to surgical intervention, there was a robust asymmetry between the vocal folds, as characterized by a greater spectral power in the contralateral vocal fold in comparison with the pathologic vocal fold (Figure 2). Following surgical intervention, a significant increase in the total spectral power in the treated pathologic vocal fold was present (P = 0.03). Moreover, a marked increase in spectral power was also noted in the untreated contralateral vocal fold, which was trending toward significance (P = 0.07; Figure 2). Spectral shape Prior to surgical intervention, the contralateral vocal fold vibratory spectrum presented with greater power in comparison with the treated vocal fold vibratory spectrum. This vibratory pattern was seen across all spectral peaks (H1, H2, and H3; Figure 3). Following surgical intervention, the two vocal folds presented with varying changes in spectral power. In the treated vocal fold, a significant increase in spectral power (P = 0.025) was noted at H1, with little change in spectral power of the higher harmonics following surgery. By contrast, the contralateral vocal fold presented with a slight increase in spectral power across both H1 and the higher harmonics. Spectral power across the diagnoses Despite the improvements in vocal fold vibration following surgical intervention, the degree of vibratory change in each vocal fold was influenced by the diagnosis (Figure 4). In subjects with a vocal fold polyp, improved vibratory power was noted across both vocal folds. The subject with unilateral vocal fold paralysis was combined into subjects with vocal fold paresis. In subjects
VHI-10 Significant improvements in the VHI-10 scores (P = 0.005; average decrease in the VHI-10 score = 9.06 points) were noted following surgical intervention (Figure 5). The vocal fold polyp subjects presented with a mild handicap prior to surgical intervention. Following surgical intervention, the VHI-10 scores are within the normal range. The patients with vocal fold paresis or paralysis and scar presented with a moderate-severe handicap prior to surgical intervention. Although improvements were noted following surgical intervention, the VHI scores continue to be markedly handicapped in comparison with normal values. Comparison with normal subjects Across all subjects, surgical intervention improved spectral power. However, in comparison with normal vibratory spectrum, the postsurgical spectral power continues to decrease (Figure 4).
FIGURE 4. Mean spectral values of F0, H2, and H3 prior to and following surgical intervention in the treated pathologic vocal fold (Tx VF) and the untreated contralateral vocal fold (contra VF).
ARTICLE IN PRESS 4
FIGURE 5. Voice Handicap Index prior to surgery and following surgery across the three diagnoses.
Postsurgical spectral power values were lower than normal subjects across all spectral peaks, with the greatest difference in H1 and H2. Such observation was further noted in the VHI-10 scores, where VHI-10 is overall greater than normative values (Figure 5). DISCUSSION HSV with spectral analysis allows examination of the precise cycle-to-cycle vibratory changes in both vocal folds following surgical intervention. To our knowledge, this is the first study to utilize HSV to examine quantitative vibratory changes specific to the treated pathologic vocal fold, and also for the contralateral vocal fold following unilateral laryngeal surgical intervention. Our results, albeit preliminary, demonstrate the following important findings. First, although overall improvements in vocal fold vibration have been reported in the literature,9,11–13,19,20 our study further demonstrated that improvements differed between the treated and the contralateral vocal fold. The treated vocal fold presented with a significant increase in spectral power following surgery. This change was most significant at the fundamental frequency, with limited change of spectral power at the higher harmonics. On the other hand, the contralateral vocal fold also presented with improvements in spectral power, which was trending toward significance. The postoperative spectral changes were characterized by an overall increase in spectral power at H1, as well as the higher harmonics. Gauffin and Sundberg16 reported that the spectral energy of the first harmonic is associated with the degree of vocal fold excursion, whereas the higher harmonic energies are associated with the discontinuity that occurs with vocal fold impact. Therefore, vibratory changes observed in the treated vocal fold suggested improvement of lateral excursion, with little change in the discontinuity of impact. By contrast, the contralateral vocal fold demonstrated improvements in both lateral excursion as well as the discontinuity of impact. This study provides direct physiological evidence that vibratory power increases in both vocal folds when surgery is limited to one vocal fold. Furthermore, there appears to be a specific change in vibratory pattern corresponding to each vocal fold following unilateral vocal fold
Journal of Voice, Vol. ■■, No. ■■, 2017
surgery. This could be due to the improved closure by removal of the mass or by injection, resulting in improved aerodynamic parameters that allow better vibratory function in the unoperated side. Although improved spectral power was noted across all subjects, changes in vibratory spectrum were also influenced by the laryngeal pathology. In subjects with a benign vocal fold lesion, we observed a good restoration of vibratory power in both treated and contralateral vocal folds following resection of the lesion. In subjects with a unilateral vocal fold paresis or paralysis, we demonstrated a marked increase in the treated vocal fold, as compared with the contralateral vocal fold. In subjects with scarring of the vocal fold, we saw mild improvements of the treated vocal fold. Interestingly, the majority of the improvement was seen in the contralateral vocal fold. Such findings suggest improvements in voice may result from improving vibration of the contralateral vocal fold as opposed to the treated vocal fold in patients with vocal fold scar. These findings further highlight the importance of considering the pathologic vocal fold, and equally important, the contralateral vocal fold during assessment and management of vocal pathology. When surgical and behavioral intervention resulted in little vibratory gains in the pathologic vocal fold, treatment in both vocal folds may be an alternative. Improvements in vocal fold vibration were greater in pathology involving the most superficial layers of the vocal fold in comparison with deeper layers. Although vocal fold polyp involves only the superficial layer with little impact on mucosal wave, vocal fold scar involves deep in the vocal fold with impact on stiffness and mucosal wave. The reduced pliability may limit improvement in displacement of the vocal fold as well as aerodynamic parameters. It should be noted that subjects with benign vocal fold lesions were reported to have a lower VHI-10 score than subjects with vocal fold scar. Given the preliminary nature of this study, further study with greater participants is currently underway to elucidate the precise vibratory changes corresponding to specific vocal fold pathology. Finally, despite significant improvements in vibratory power following unilateral vocal fold surgery, postoperative vibratory spectrum continues to be reduced in comparison with normal vibratory spectrum. This was evident across all spectral peaks at the fundamental as well as across higher harmonics. The findings were further corroborated with distinct differences in the VHI-10 score between normal and treated pathologic voices following surgical intervention. Previous findings have found that following surgical intervention, vocal fold vibration continues to be reduced in comparison with normal vibration.9 Our study further expanded previous studies in detailing that both the pathologic and the contralateral vocal folds vibrate at a decreased manner as compared with normal vocal fold vibration. This study was not able to study other physiological variables, such as improved subglottic pressure and improved Alternating current/direct current (AC/DC) flow characteristics after phonosurgery. Improved vocal tract conditions by removal of masses, improved closure, and softening of scar would be expected to improve vibratory conditions for both vocal folds and vocal tract aerodynamics, thereby possibly explaining the
ARTICLE IN PRESS Wenli Chen, et al
Vocal Fold Vibration Following Surgical Intervention
better vibration noted in the contralateral vocal fold in all parameters measured. CONCLUSION HSV provides cycle-to-cycle quantitative information regarding vibration of the vocal folds. To our knowledge, this is the first study to examine quantitative vibratory changes specific to the treated pathologic vocal fold and also to the contralateral vocal fold following surgical intervention. Despite improvements in vibratory power, there continues to be a distinct difference in vibration in comparison with normal vocal folds. This highlights the importance of precise postsurgical assessments in guiding subsequent surgical management and voice rehabilitation. REFERENCES 1. Hansworth DW. High-speed motion pictures of the human vocal cords. Bell Lab Record. 1940;18:203–208. 2. Moore GP, White FD, Von Leden H. Ultra-high speed photography in laryngeal physiology. J Speech Hear Disord. 1962;27:165–171. 3. Hertegard S, Larsson H, Wittenberg T. High-speed imaging: applications and development. Logoped Phoniatr Vocol. 2003;28:133–139. 4. Kitzing P. Stroboscopy—a pertinent laryngeal examination. J Otolaryngol. 1985;14:151–157. 5. Ahmad K, Yan Y, Bless D. Vocal-fold vibratory characteristics in normal female speakers from high-speed digital imaging. J Voice. 2012;26:239–253. 6. Yamauchi A, Imagawa H, Yokonishi H, et al. Evaluation of vocal fold vibration with an assessment form for high-speed digital imaging: comparative study between healthy young and elderly subjects. J Voice. 2012;26:742–750. 7. Chen W, Woo P, Murry T. Spectral analysis of digital kymography in normal adult vocal fold vibration. J Voice. 2014;28:356–361.
5
8. Yan Y, Ahmad K, Kunduk M, et al. Analysis of vocal-fold vibrations from high-speed laryngeal images using Hilbert transform-based methodology. J Voice. 2005;19:161–175. 9. Chen W, Woo P, Murry T. Vocal fold vibratory changes following surgical intervention. J Voice. 2016;30:224–227. 10. Yan Y, Damrose E, Bless D. Functional analysis of voice using simultaneous high-speed imaging and acoustic recordings. J Voice. 2007;21:604–616. 11. Kunduk M, Dollinger M, McWhorter AJ, et al. Vocal fold vibratory behavior changes following surgical treatment of polyps investigated with high-speed videoendoscopy and phonovibrography. Ann Otol Rhinol Laryngol. 2012;121:355–363. 12. Bonilha H, Deliyski D, Whiteside J, et al. Vocal fold phase asymmetries in patients with voice disorders: a study across visualization techniques. Am J Speech Lang Pathol. 2012;21:3–15. 13. Kimura M, Nito T, Imagawa H, et al. Collagen injection for correcting vocal fold asymmetry: high-speed imaging. Ann Otol Rhinol Laryngol. 2010;119:359–368. 14. Wittenberg T, Tigges M, Mergell P, et al. Functional imaging of vocal fold vibration: digital multislice high-speed kymography. J Voice. 2000;14:422– 442. 15. Svec JG, Schutte HK. Videokymography: high-speed line scanning of vocal fold vibration. J Voice. 1996;10:201–205. 16. Gauffin J, Sundberg J. Spectral correlates of glottal voice source waveform characteristics. J Speech Hear Res. 1989;32:556–565. 17. Bonilha H, Deliyski D. Period and glottal width irregularities in vocally normal speakers. J Voice. 2008;22:699–708. 18. Patel R, Dixon A, Richmond A, et al. Pediatric high speed digital imaging of vocal fold vibration: a pilot normative study of glottal closure and phase closure. Int J Pediatr Otorhinolaryngol. 2012;76:954–959. 19. Noordzij JP, Woo P. Glottal area waveform analysis of benign vocal fold lesions before and after surgery. Ann Otol Rhinol Laryngol. 2000;109:441– 446. 20. Pearl AW, Woo P, Ostrowski R, et al. A preliminary report on micronized AlloDerm injection laryngoplasty. Laryngoscope. 2002;112:990–996.
INTRODUCTION Postoperative assessment of vocal fold vibration is imperative from a surgical perspective as well as a rehabilitative perspective. It determines surgical success, candidacy for voice therapy, and whether further medical management is needed. Current postoperative assessments of vocal fold vibration consist of subjective analysis utilizing videostroboscopy, self-assessment tools, or audio or aerodynamic instrumental analysis. In many cases, the patient is left to making a simple self-perception judgment of “better, worse, or the same.” High-speed videoendoscopy (HSV) captures data at more than 2000 frames per second, which allows direct visualization of the entire cycle-to-cycle vibratory motion of the left and right vocal fold in normal speaking adults1–8 and in disordered phonation.9–15 The use of HSV has provided significant utility in understanding the etiology of disorders on vocal fold vibration, as well as provided close examination of aperiodic vibrations that are difficult to capture on standard stroboscopy measures.9–15 Digital kymography uses HSV to examine the precise vibratory characteristics of each vocal fold at selected locations along the vocal folds.7,9,14,15 The resulting values can be analyzed as a vibratory spectrum.7,9 In the vibratory spectrum, spectral power in the fundamental frequency (F0 = H1) has been associated with the degree of vocal fold excursion, whereas the energy of the higher harmonics has been associated with the discontinuity that occurs with vocal fold impact.9,16 Therefore, digital kymography spectrum is a useful tool to objectively quantify precise vocal
Accepted for publication February 14, 2017. A version of the paper was presented at the 2015 Fall Voice Conference in Pittsburgh, PA on October 17, 2015. From the *Department of Otorhinolaryngology, Mayo Clinic, Phoenix, Arizona; †Department of Otolaryngology Head and Neck Surgery, Icahn School of Medicine, New York, New York; and the ‡Department of Otolaryngology Head and Neck Surgery, Loma Linda University, Loma Linda, California. Address correspondence and reprint requests to Wenli Chen, Mayo Clinic, 5779 E Mayo Blvd, Phoenix, AZ 85054. E-mail: [email protected] Journal of Voice, Vol. ■■, No. ■■, pp. ■■-■■ 0892-1997 © 2017 The Voice Foundation. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jvoice.2017.02.006
fold vibration and offers an ideal methodology to examine the specific changes in vibratory behavior of the vocal folds following surgical intervention.9 Quantification of vibration in each vocal fold using HSV is now practical. Previous studies using HSV demonstrated improvements in vocal fold vibration following medical intervention. Kunduk et al11 have shown improved amplitude and symmetry of vocal fold vibration following surgical resection of vocal fold polyp at 1 month and 3 months postsurgery. Kimura et al13 demonstrated improvements in vocal vibratory patterns toward symmetry in patients with unilateral vocal fold paralysis following collagen injection. Chen et al9 have shown improved vibratory amplitude and symmetry in three types of vocal fold pathologies following surgical treatment, with the most remarkable change noted at the mid-membranous region of the vocal fold. Although these studies unequivocally demonstrate improved vibratory motion in both vocal folds, little is known about the specific vibratory changes in each vocal fold. Understanding changes in the treated as well as the contralateral unoperated vocal fold may assist in surgical management and offer insight as to the basis of remaining dysphonia or the lack of it. Therefore, the goal of this preliminary study was to examine direct vocal fold vibratory changes in the treated pathologic vocal fold and the contralateral vocal fold following surgical intervention. METHODS Three subjects with identifiable mid-vocal fold benign lesions, two with unilateral vocal fold paresis, one with unilateral vocal fold paralysis, and three with vocal fold scar were studied prior to and following surgical intervention. Vocal pathology was diagnosed by a board-certified otolaryngologist. These diagnostic types were intentionally selected to represent varying levels of vocal fold pathology (refer to Table 1 for laryngeal etiology, age, diagnosis, and the type of surgical intervention). Two male and one female normal adult subjects were also recorded for comparison. HSV was obtained using standard videostroboscopy procedures.
ARTICLE IN PRESS 2
Journal of Voice, Vol. ■■, No. ■■, 2017
TABLE 1. Information on Pathologic Group, Age, Diagnosis, and Type of Surgical Intervention Received Subject M1 M2 M3 S1 S2 S3 P1 P2 P3
Age
Diagnosis
Diagnostic Information
Surgical Intervention
42 45 47 85 71 68 45 49 67
Polyp Polyp Polyp Scar Scar Scar Paralysis Paresis Paresis
L VF polyp R VF polyp L VF polyp R VF hyperkeratosis L VF scar with recurrent keratosis L VF scar with R VF bowing R VF paralysis with R sulcus vocalis L VF paresis L VF paresis
Polyp excision using cold instruments Polyp excision using cold instruments KTP laser KTP laser KTP laser KTP laser Injection laryngoplasty Injection laryngoplasty Injection laryngoplasty
Abbreviation: KTP, potassium titanyl phosphate; L, left; R, right; VF, vocal fold.
High-speed video recording HSV was recorded with the Kay Elemetrics High-Speed Digital Imaging (HSDI) system (KayPENTAX Photron Motion, Montvale, NJ, USA), which consisted of a 90° rigid endoscope (Model 9100) and a 300-Watt Xenon light source. The HSDI system captured grayscale images at a rate of 2000 frames per second, with a spatial resolution of 256 × 120 pixels rotated to a vertical position. Videostrobolaryngoscopy was conducted utilizing a rigid endoscope, with the subject sustaining /i/ at a comfortable pitch and loudness. A contact microphone was attached at the neck to monitor pitch and a second microphone was placed 6 inches from the lips to monitor intensity. The HSV samples were obtained when the examiner observed a clear and full view of the larynx. Six continuous 2-second tokens of the vowel were recorded. The three best tokens with a clear and full view of the vocal folds were saved onto a hard drive for analysis. All subjects tolerated the data collection procedure without any difficulties. Data analysis Kymography image processing HSV images were preprocessed with video editing software (VirtualDub v.1.9.11 (virtualdub.org)). Brightness and contrast of the glottis and vocal folds were adjusted for optimal edge detection. Image rotation was implemented, as needed, to ensure vertical alignment of the image. A 400- to 500-frame video segment that captured a full view of the vocal folds with minimal movement of the subject was extracted from the recorded HSV samples. Kay’s
Image Processing Software (KIPS, Model 9181) was used to generate the kymogram. Digital kymograph was created by placing a transverse line across the mid-membranous region of the glottis (Figure 1A and B), which has been reported to be the area of maximal vocal fold contact.17 Edge detection was applied to trace the vocal fold edges (Figure 1C). Kymograph analysis of the vibratory samples is dependent on the delineation of the vocal fold edge from HSV.18 Therefore, manual corrections function was utilized, as needed, to ensure correct tracing of the vocal fold edges. Kymograph edge analysis function was subsequently applied on the kymogram. The resulting values were kymograph edge data, which showed the coordinate values of the left and right edges of the vocal fold across time (Figure 1D). Fourier transform function was subsequently applied. This resulted in a spectrum ranging from 0 Hz to 1000 Hz for the left and right edges of the vocal folds (Figure 1E). Spectral data analyses Three values were extracted from the left and the right vocal fold vibratory spectrum for quantitative analysis: the peak power values of the fundamental (F0 = H1), second harmonic (H2), and third harmonic (H3). Twenty-five percent of the postprocessed 500frame HSV tokens were randomly selected and reanalyzed by the same experimenter to evaluate the error of measurement. Comparison of the H1, H2, and H3 values between the original and reanalyzed sample yielded a reliability of 87%. The following vocal fold vibratory parameters were extracted at baseline and postintervention: (1) total spectral power,
FIGURE 1. Methods to obtain spectral analysis of digital kymography in subject P2 (see text for descriptions).
ARTICLE IN PRESS Wenli Chen, et al
Vocal Fold Vibration Following Surgical Intervention
3
FIGURE 2. Mean values of total spectral power prior to and following surgical intervention in the treated pathologic vocal fold (Tx VF) and the untreated contralateral vocal fold (contra VF).
FIGURE 3. Mean spectral values of F0, H2, and H3 prior to and following surgical intervention in the treated pathologic vocal fold (treated VF) and the untreated contralateral vocal fold (contra VF).
(2) spectrum shape, and (3) the Voice Handicap Index-10 (VHI10). The total spectral power was determined by calculating the sum of the spectral peak values of H1, H2, and H3 altogether. Spectrum shape was determined by tabulating the spectral power at H1, H2, and H3 individually. Lastly, VHI-10 was administered to examine subjects’ perceptions of their voice handicap. Statistical analysis was performed with t test using the JMP software (JMP Pro 10, SAS Institute, Cary, NC, USA).
with a unilateral vocal fold paresis or paralysis, a marked increase in spectral power was noted in the treated vocal fold as compared with the contralateral vocal fold. Subjects with a vocal fold scar presented with a marked increase in the spectral power of the contralateral vocal fold as compared with the treated vocal fold. The greatest change in the total spectral power was noted in the polyp group. The least change was noted in the scar group.
RESULTS Total spectral power Prior to surgical intervention, there was a robust asymmetry between the vocal folds, as characterized by a greater spectral power in the contralateral vocal fold in comparison with the pathologic vocal fold (Figure 2). Following surgical intervention, a significant increase in the total spectral power in the treated pathologic vocal fold was present (P = 0.03). Moreover, a marked increase in spectral power was also noted in the untreated contralateral vocal fold, which was trending toward significance (P = 0.07; Figure 2). Spectral shape Prior to surgical intervention, the contralateral vocal fold vibratory spectrum presented with greater power in comparison with the treated vocal fold vibratory spectrum. This vibratory pattern was seen across all spectral peaks (H1, H2, and H3; Figure 3). Following surgical intervention, the two vocal folds presented with varying changes in spectral power. In the treated vocal fold, a significant increase in spectral power (P = 0.025) was noted at H1, with little change in spectral power of the higher harmonics following surgery. By contrast, the contralateral vocal fold presented with a slight increase in spectral power across both H1 and the higher harmonics. Spectral power across the diagnoses Despite the improvements in vocal fold vibration following surgical intervention, the degree of vibratory change in each vocal fold was influenced by the diagnosis (Figure 4). In subjects with a vocal fold polyp, improved vibratory power was noted across both vocal folds. The subject with unilateral vocal fold paralysis was combined into subjects with vocal fold paresis. In subjects
VHI-10 Significant improvements in the VHI-10 scores (P = 0.005; average decrease in the VHI-10 score = 9.06 points) were noted following surgical intervention (Figure 5). The vocal fold polyp subjects presented with a mild handicap prior to surgical intervention. Following surgical intervention, the VHI-10 scores are within the normal range. The patients with vocal fold paresis or paralysis and scar presented with a moderate-severe handicap prior to surgical intervention. Although improvements were noted following surgical intervention, the VHI scores continue to be markedly handicapped in comparison with normal values. Comparison with normal subjects Across all subjects, surgical intervention improved spectral power. However, in comparison with normal vibratory spectrum, the postsurgical spectral power continues to decrease (Figure 4).
FIGURE 4. Mean spectral values of F0, H2, and H3 prior to and following surgical intervention in the treated pathologic vocal fold (Tx VF) and the untreated contralateral vocal fold (contra VF).
ARTICLE IN PRESS 4
FIGURE 5. Voice Handicap Index prior to surgery and following surgery across the three diagnoses.
Postsurgical spectral power values were lower than normal subjects across all spectral peaks, with the greatest difference in H1 and H2. Such observation was further noted in the VHI-10 scores, where VHI-10 is overall greater than normative values (Figure 5). DISCUSSION HSV with spectral analysis allows examination of the precise cycle-to-cycle vibratory changes in both vocal folds following surgical intervention. To our knowledge, this is the first study to utilize HSV to examine quantitative vibratory changes specific to the treated pathologic vocal fold, and also for the contralateral vocal fold following unilateral laryngeal surgical intervention. Our results, albeit preliminary, demonstrate the following important findings. First, although overall improvements in vocal fold vibration have been reported in the literature,9,11–13,19,20 our study further demonstrated that improvements differed between the treated and the contralateral vocal fold. The treated vocal fold presented with a significant increase in spectral power following surgery. This change was most significant at the fundamental frequency, with limited change of spectral power at the higher harmonics. On the other hand, the contralateral vocal fold also presented with improvements in spectral power, which was trending toward significance. The postoperative spectral changes were characterized by an overall increase in spectral power at H1, as well as the higher harmonics. Gauffin and Sundberg16 reported that the spectral energy of the first harmonic is associated with the degree of vocal fold excursion, whereas the higher harmonic energies are associated with the discontinuity that occurs with vocal fold impact. Therefore, vibratory changes observed in the treated vocal fold suggested improvement of lateral excursion, with little change in the discontinuity of impact. By contrast, the contralateral vocal fold demonstrated improvements in both lateral excursion as well as the discontinuity of impact. This study provides direct physiological evidence that vibratory power increases in both vocal folds when surgery is limited to one vocal fold. Furthermore, there appears to be a specific change in vibratory pattern corresponding to each vocal fold following unilateral vocal fold
Journal of Voice, Vol. ■■, No. ■■, 2017
surgery. This could be due to the improved closure by removal of the mass or by injection, resulting in improved aerodynamic parameters that allow better vibratory function in the unoperated side. Although improved spectral power was noted across all subjects, changes in vibratory spectrum were also influenced by the laryngeal pathology. In subjects with a benign vocal fold lesion, we observed a good restoration of vibratory power in both treated and contralateral vocal folds following resection of the lesion. In subjects with a unilateral vocal fold paresis or paralysis, we demonstrated a marked increase in the treated vocal fold, as compared with the contralateral vocal fold. In subjects with scarring of the vocal fold, we saw mild improvements of the treated vocal fold. Interestingly, the majority of the improvement was seen in the contralateral vocal fold. Such findings suggest improvements in voice may result from improving vibration of the contralateral vocal fold as opposed to the treated vocal fold in patients with vocal fold scar. These findings further highlight the importance of considering the pathologic vocal fold, and equally important, the contralateral vocal fold during assessment and management of vocal pathology. When surgical and behavioral intervention resulted in little vibratory gains in the pathologic vocal fold, treatment in both vocal folds may be an alternative. Improvements in vocal fold vibration were greater in pathology involving the most superficial layers of the vocal fold in comparison with deeper layers. Although vocal fold polyp involves only the superficial layer with little impact on mucosal wave, vocal fold scar involves deep in the vocal fold with impact on stiffness and mucosal wave. The reduced pliability may limit improvement in displacement of the vocal fold as well as aerodynamic parameters. It should be noted that subjects with benign vocal fold lesions were reported to have a lower VHI-10 score than subjects with vocal fold scar. Given the preliminary nature of this study, further study with greater participants is currently underway to elucidate the precise vibratory changes corresponding to specific vocal fold pathology. Finally, despite significant improvements in vibratory power following unilateral vocal fold surgery, postoperative vibratory spectrum continues to be reduced in comparison with normal vibratory spectrum. This was evident across all spectral peaks at the fundamental as well as across higher harmonics. The findings were further corroborated with distinct differences in the VHI-10 score between normal and treated pathologic voices following surgical intervention. Previous findings have found that following surgical intervention, vocal fold vibration continues to be reduced in comparison with normal vibration.9 Our study further expanded previous studies in detailing that both the pathologic and the contralateral vocal folds vibrate at a decreased manner as compared with normal vocal fold vibration. This study was not able to study other physiological variables, such as improved subglottic pressure and improved Alternating current/direct current (AC/DC) flow characteristics after phonosurgery. Improved vocal tract conditions by removal of masses, improved closure, and softening of scar would be expected to improve vibratory conditions for both vocal folds and vocal tract aerodynamics, thereby possibly explaining the
ARTICLE IN PRESS Wenli Chen, et al
Vocal Fold Vibration Following Surgical Intervention
better vibration noted in the contralateral vocal fold in all parameters measured. CONCLUSION HSV provides cycle-to-cycle quantitative information regarding vibration of the vocal folds. To our knowledge, this is the first study to examine quantitative vibratory changes specific to the treated pathologic vocal fold and also to the contralateral vocal fold following surgical intervention. Despite improvements in vibratory power, there continues to be a distinct difference in vibration in comparison with normal vocal folds. This highlights the importance of precise postsurgical assessments in guiding subsequent surgical management and voice rehabilitation. REFERENCES 1. Hansworth DW. High-speed motion pictures of the human vocal cords. Bell Lab Record. 1940;18:203–208. 2. Moore GP, White FD, Von Leden H. Ultra-high speed photography in laryngeal physiology. J Speech Hear Disord. 1962;27:165–171. 3. Hertegard S, Larsson H, Wittenberg T. High-speed imaging: applications and development. Logoped Phoniatr Vocol. 2003;28:133–139. 4. Kitzing P. Stroboscopy—a pertinent laryngeal examination. J Otolaryngol. 1985;14:151–157. 5. Ahmad K, Yan Y, Bless D. Vocal-fold vibratory characteristics in normal female speakers from high-speed digital imaging. J Voice. 2012;26:239–253. 6. Yamauchi A, Imagawa H, Yokonishi H, et al. Evaluation of vocal fold vibration with an assessment form for high-speed digital imaging: comparative study between healthy young and elderly subjects. J Voice. 2012;26:742–750. 7. Chen W, Woo P, Murry T. Spectral analysis of digital kymography in normal adult vocal fold vibration. J Voice. 2014;28:356–361.
5
8. Yan Y, Ahmad K, Kunduk M, et al. Analysis of vocal-fold vibrations from high-speed laryngeal images using Hilbert transform-based methodology. J Voice. 2005;19:161–175. 9. Chen W, Woo P, Murry T. Vocal fold vibratory changes following surgical intervention. J Voice. 2016;30:224–227. 10. Yan Y, Damrose E, Bless D. Functional analysis of voice using simultaneous high-speed imaging and acoustic recordings. J Voice. 2007;21:604–616. 11. Kunduk M, Dollinger M, McWhorter AJ, et al. Vocal fold vibratory behavior changes following surgical treatment of polyps investigated with high-speed videoendoscopy and phonovibrography. Ann Otol Rhinol Laryngol. 2012;121:355–363. 12. Bonilha H, Deliyski D, Whiteside J, et al. Vocal fold phase asymmetries in patients with voice disorders: a study across visualization techniques. Am J Speech Lang Pathol. 2012;21:3–15. 13. Kimura M, Nito T, Imagawa H, et al. Collagen injection for correcting vocal fold asymmetry: high-speed imaging. Ann Otol Rhinol Laryngol. 2010;119:359–368. 14. Wittenberg T, Tigges M, Mergell P, et al. Functional imaging of vocal fold vibration: digital multislice high-speed kymography. J Voice. 2000;14:422– 442. 15. Svec JG, Schutte HK. Videokymography: high-speed line scanning of vocal fold vibration. J Voice. 1996;10:201–205. 16. Gauffin J, Sundberg J. Spectral correlates of glottal voice source waveform characteristics. J Speech Hear Res. 1989;32:556–565. 17. Bonilha H, Deliyski D. Period and glottal width irregularities in vocally normal speakers. J Voice. 2008;22:699–708. 18. Patel R, Dixon A, Richmond A, et al. Pediatric high speed digital imaging of vocal fold vibration: a pilot normative study of glottal closure and phase closure. Int J Pediatr Otorhinolaryngol. 2012;76:954–959. 19. Noordzij JP, Woo P. Glottal area waveform analysis of benign vocal fold lesions before and after surgery. Ann Otol Rhinol Laryngol. 2000;109:441– 446. 20. Pearl AW, Woo P, Ostrowski R, et al. A preliminary report on micronized AlloDerm injection laryngoplasty. Laryngoscope. 2002;112:990–996.