Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Dec 1.
Published in final edited form as: Laryngoscope. 2023 Jun 19;133(12):3482–3491. doi: 10.1002/lary.30830

Extrinsic Laryngeal Muscle Tension in Primary Muscle Tension Dysphonia with Shear Wave Elastography

Adrianna C Shembel 1,2, Robert A Morrison 2, David T Fetzer 3, Amber Patterson-Lachowicz 3, Sarah McDowell 2, Julianna C Comstock Smeltzer 2, Ted Mau 1
PMCID: PMC10728340  NIHMSID: NIHMS1909131  PMID: 37334857

Abstract

Objectives:

It has been assumed that patients with primary muscle tension dysphonia (pMTD) have more extrinsic laryngeal muscle (ELM) tension, but tools to study this phenomenon lack. Shear wave elastography (SWE) is a potential method to address these shortcomings. The objectives of this study were to apply SWE to the ELMs; compare SWE measures to standard clinical metrics; and determine group differences in pMTD and typical voice users before and after vocal load.

Methods:

SWE measurements of the ELMs from ultrasound examinations of the anterior neck, supraglottic compression severities from laryngoscopic images, cepstral peak prominences (CPP) from voice recordings, and self-perceptual ratings of vocal effort and discomfort were obtained in voice users with (N=30) and without (N=35) pMTD, before and after a vocal load challenge.

Results:

ELM tension significantly increased from rest-to-voiced conditions in both groups. However, the groups were similar in their ELM stiffness levels at SWE at baseline, during vocalization, and post-vocal load. Levels of vocal effort and discomfort and supraglottic compression was significantly higher and CPP significantly lower in the pMTD group. Vocal load had a significant effect on vocal effort and discomfort but not on laryngeal or acoustic patterns.

Conclusion:

SWE can be used to quantify ELM tension with voicing. Although the pMTD group reported significantly higher levels of vocal effort and vocal tract discomfort and, on average, exhibited significantly more severe supraglottic compression and lower CPP values, there were no group differences on levels of ELM tension using SWE.

Keywords: vocal hyperfunction, ultrasound, elastography, vocal effort, vocal tract discomfort, muscle tension dysphonia, shear wave elastography, extrinsic laryngeal muscle

Lay Summary:

Patients with muscle tension dysphonia are thought to have tension in the muscles around their voice box. However, this study using an ultrasound method to measure tension in these muscles (shear wave elastography) found no group differences.

INTRODUCTION

Clinical wisdom is that symptoms of increased vocal effort, vocal fatigue, vocal tract and neck discomfort, and odynophonia in primary muscle tension dysphonia (pMTD) are the result of excessive extrinsic laryngeal muscle (ELM) tension.1–3 Occupational voice users are especially at risk. However, the relationship between these voice symptoms and ELM tension has been challenging to deduce due to the lack of well-vetted physiological metrics to study muscle tension in the ELMs. Current approaches for ELM assessment in pMTD include perilaryngeal palpation and electromyography, which have poor reliability and specificity.4,5 These subjective approaches can provide information on ELM tone at best, but do not inform biomechanical properties or contraction heterogeneity of individual muscles within the extrinsic laryngeal musculoskeletal system that affect force generation (tension) and range of motion (hyperfunction).

Standard instrumental voice assessments (acoustics, aerodynamics) can confirm the presence and quantify the severity of a disordered voice. However, these instrumental voice metrics have poor diagnostic specificity for the pMTD population6 and do not inform the role of the ELMs in pMTD pathophysiology. Furthermore, these voice instrumentation methods may not always be relevant to the pMTD population since it is not uncommon for patients with pMTD to have complaints with the feel of their voice (e.g., effort, discomfort) while the sound of their voice is within normal limits.6,7 Physiological methods that can quantify ELM tension and study relationships between ELM tension and patient complaints of vocal effort, fatigue, and discomfort central to patients with pMTD—thought to result from these muscle tension patterns—are needed.

One approach to overcome these methodological shortcomings is the use an ultrasound technique—shear wave elastography (SWE)—to measure ELM tension. SWE ultrasound uses acoustic force impulse technology to estimate tissue stiffness—or intramuscular tension—by measuring the velocities of shear waves (Figure 1). SWE has previously been validated in and can be applied to any tissue that is elastic, like skeletal muscle. There are three factors that are essential for accurate and reliable SWE capture. The first is that SWE is sensitive to applied transducer pressure, requiring a skilled sonographer to keep pressures across trials, patients, and days consistent. The second is that more superficial muscles, like the muscles of the anterior neck, are more reliable than deeper muscles, the latter of which may have more variability across trials and the potential for greater shear wave attenuation effects.8 The third is that large muscles at maximum voluntary contraction, like in the quadriceps, can decrease the accuracy of shear wave velocities,8 which is less of an issue in the smaller ELMs, especially when involved in speech production—a submaximal task. As such, knowledge of the ultrasound equipment, anisotropic physiological properties of the target muscle, identification of ideal muscles for this technique, and a skilled sonographer are all essential for reliable and valid measures.9

Figure 1. Shear Wave Elastography Schematic and Representative Images.

Figure 1.

Open in a new tab

(A) Schematic of acoustic push pulses generated by the transducer into the target tissue. This acoustic pulse focally compresses the underlying tissue and induces shear waves that propagate laterally. The higher the propagation velocity of the shear wave, the higher the degree of tissue stiffness—or intramuscular tension—based on Young’s Modulus. (B) Grayscale ultrasound image of the target muscle, with three 3mm diameter circular regions of interests (ROIs) placed on the thyrohyoid muscle (1–3) and sternohyoid muscle (4–6) to obtain quantitative estimate of shear wave velocities within the ROIs. (C)-(D) Quantitative elastograms represented as a color-coded heat map relative to other surrounding tissue within a rectangular region of interest. The faster the shear wave velocities, the greater the stiffness in the target tissue, which are represented as warmer colors on an overlaid heat map (D). Conversely, the slower the velocities, the more pliable and less stiff the tissue, which is represented as cooler colors on the overlaid heat map (C). Panel (C) is a representative image of a control subject at rest. Panel (D) is a representative image of a control subject during increased pitch (C5; 537 Hz) when the extrinsic muscles are known to be most active.45–47 The thyroid cartilage is “hot” (i.e., red) in this photo because it is not only very stiff tissue but also due to acoustic reverberations against stiff tissue during vocalization that not affect the muscle. Schematic partially created with BioRender.com.

When performed correctly, quantitative analysis using SWE has been shown to be reliable with low variability across trials, within different ranges of interest within and across different skeletal muscle groups, across users, and with inter-day repeatability.8–17 SWE has also successfully been used to study group differences between patients with Duchenne muscular dystrophy (a muscle disorder of increased muscle stiffness) and healthy controls,18 muscle stiffness after exercise-induced muscle damage,19 changes in muscle tension after application of Kinesio (deloading) tape to the skin directly over the muscle (rectus femoris),20 the effects of massage on reducing muscle stiffness,21 and sex differences in muscle fatigue after a peripheral neuromuscular fatigability task of the knee extensor.10

Considering that ELM tension is thought to be ubiquitous to patients with pMTD, it would be expected that shear wave velocities in the ELMs would be higher at baseline (rest) and even higher with phonation and increased vocal demands in those with pMTD compared to typical voice users. Second, considering that ELM tension is thought to underlie symptoms of vocal effort and discomfort, it would be expected that those with higher vocal effort and discomfort would have higher ELM tension. Third, considering ELM tension is attributed to aberrant vocal quality and correlates with supraglottic laryngeal patterns in patients with pMTD, then patients with these clinical patterns should also exhibit higher ELM tension.

The first objective of this study was to compare the level of muscle tension in the ELMs between occupational voice users with pMTD and typical occupational voice users without vocal dysfunction using SWE. The second objective was to compare SWE measures in the same two groups to standard self-perceptual and laryngoscopic metrics. The third objective was to determine group differences in SWE measures before and after an increased vocal demand (vocal load task). This third objective was driven by the classic presentation of increased voice use or heavy vocal demands that often precipitate the onset and perpetuation of pMTD, and the high prevalence of pMTD in occupational voice users who rely on their voices for their profession.2 The overall hypothesis was that occupational voice users with pMTD had more ELM tension than those without vocal hyperfunction.

METHODS

The study design involved a 2 × 2 mixed methods design with group (pMTD, control) and time point (pre-vocal load, post-vocal load) as independent variables. The shear wave velocities of the four ELMs (anterior digastric, geniohyoid, thyrohyoid, and sternohyoid), vocal effort (100 mm visual analog scale), vocal tract discomfort (Vocal Tract Discomfort Scale22,23), cepstral peak prominence (acoustic voice sample), and mediolateral supraglottic compression (laryngoscopy) were outcome variables. All variables were measured before and after the half-hour vocal load task.

Participants

Occupational voice users were recruited for this study because we were interested in understanding ELM contributions specific to pathological reports of vocal effort, fatigue, and discomfort separate from ELM patterns distinct from heavy voice use—a common factor in patients with pMTD.24–27 In other words, if there is ELM tension, and many patients with pMTD are heavy voice users, it would be unclear how much of the ELM tension is due to a voice disorder and how much is the result of occupational voice demands. Sixty-five individuals (ages 18–65 years; 78% women) in vocally demanding occupations (e.g., educators, performers), with (n = 30 experimental) and without (n = 35 control) a formal diagnosis of pMTD, were recruited for the study. A “vocally demanding” profession was defined as at least 2 hours of continuous voice use over an 8-hour day for two or more consecutive days. Participants in the pMTD group consisted of patients seen for their initial assessment at the University of Texas at Southwestern (UTSW) Center for Voice Care, who received a formal diagnosis of pMTD at their visit, and prior to receiving voice therapy. pMTD was defined by the essential features and classification criteria from the Classification Manual for Voice Disorders-1: self-described vocal impairments impacting activities of daily living for more than 3 weeks in the absence of organic pathology and presence of supraglottic constriction during phonation on laryngoscopy.28 The pMTD cohort represented typical clinical presentations of pMTD seen in a voice clinic6,7: some patients reported issues only with the feel of their voice (effort, discomfort), while others experience issue with both the sound (vocal quality) and the feel of their voice.

Since literature supports self-perceived metrics as the most reliable indicator of pMTD,6,29,30 two such indices were used as additional inclusion criteria. The goal of these indices was to ensure significant group differences in self-perceived severity of vocal impairment using the Voice Handicap Index-10 (VHI-10) and severity of vocal fatigue—a common complaint in patients with pMTD,2,31 using the Vocal Fatigue Index (VFI). Vocal effort and vocal tract discomfort were not included in the inclusion criteria because these measures were used as outcome variables. To ensure differences between groups, participants with pMTD needed to score more than 11 out of 40 on the Voice Handicap Index-10 (VHI-10)32 (based on normative values33) and more than 24 on Part 1 of the Vocal Fatigue Index (VFI-Part1).34 Nanjundeswaran et al. (2015) previously found that the 11 items under Part 1 of the Vocal Fatigue Index can distinguish pathological levels from typical vocal fatigue experienced by occupational voice users after a heavy vocal demand.34

Participants in the control group had high vocal demands but reported no vocal impairments and had no incidences or complaints of vocal dysfunction that resulted in missed work within the past six months. Control participants were required to score less than 7 out of 4033 on the VHI-10 and less than 21 on the VFI-Part1. Exclusion criteria for both groups were absence of neck trauma, surgeries, or radiation. See Table 1 for group demographics as well as descriptive statistics demonstrating significant differences between groups on vocal severity impact (VHI-10), vocal fatigue (VFI-Part1), perceptual vocal quality (100 mm visual analog scale), fundamental frequency (mean F0), and cepstral acoustic output (L/H ratio) at the start of the study protocol.

Table 1. Participant Demographics and Clinical Characteristics.

Data presented as means (standard deviations) with p-values and percentages, as appropriate. Demographics and fundamental frequencies were similar between groups, except for mean age, which was lower in the control group. Patients with pMTD scored significantly higher in their self-reported measures of vocal deficits (VHI-10) and vocal fatigue (VFI-Part1). Clinicians rated auditory-perceptual overall dysphonia severity as significantly more severe on a 100 mm visual analog scale (VAS). Acoustic characteristics (L/H Ratio) were also significantly different between groups. These clinical characteristics demonstrate significant differences in clinical presentation between occupational voice users with and without pMTD and align with common complaints of vocal deficits, fatigue, and vocal quality impacts patients with pMTD typically experience.

pMTD (n=30) Control (n=36) p-values
Demographics:
 Age 48.82 (16.98) 32.32 (13.83) p<0.0001
 Gender 77% women
23% men		 78% women
22% men		 -----------
 Race 50% White
23% Black
13% Latinx
3% Asian
 44% White
0% Black
14% Latinx
8% Asian		 -----------
Voice Indices:
 Vocal Severity Impact -- VHI-10 score (40 pts max) 23.63 (8.70) 1.89 (1.91) p<0.0001
 Vocal Fatigue -- VFI-Part1 (44 pts max) 32.20 (6.97) 9.31 (8.15) p<0.0001
Vocal Quality / Acoustic Vocal Output:
 Auditory-perceptual dysphonia severity (0–100 VAS) 39.06 (28.54) 9.00 (7.74) p<0.0001
 Fundamental frequency mean (mean F0) 188.89 (41.29) 209.31 (45.66) p=0.23
 L/H ratio (average low energy/average high energy)* −0.64 (0.41) −1.94 (0.45) p=0.03
Open in a new tab
*

Awan & Roy, 2005

Data Acquisition Procedures

Ultrasound Imaging (shear wave elastography).

A dedicated ACUSON Sequoia scanner (Siemens Healthineers, Issaquah, WA, USA) equipped with shear wave elastography capabilities, Virtual Touch Imaging and Quantification (VTIQ) software, and a linear 4–10 MHz transducer well-suited for superficial muscles were used to acquire images and SWE measurements. Author A.P.L. acquired the data. She is a sonographer with over 10 years of experience. Acoustic coupling gel was applied to the neck and the transducer placed against the skin without excessive pressure (light touch sufficient to displace air between skin and transducer), oriented parallel to the muscle fibers for the longitudinal view and in cross-section to the muscle orientation for the transverse view. The patient sat upright in an examination chair in a neutral posture without excessive neck extension. An acoustic radiation force impulse (ARFI) technique, generating a focal compressive acoustic wave, induces localized micro-meter tissue deformation and the creation of laterally propagating shear waves. A qualitative representation of shear wave speeds within a rectangular region was visualized as a color-coded heat map (elastogram) overlayed on the B-mode ultrasound (grayscale) image (c.f., Figure 1C and 1D). The simultaneous visualization of the elastogram and grayscale information enables the operator to be guided by both anatomical and physiological tissue stiffness information. Multiple 3 mm circular regions of interest (ROIs) were placed on the target ELMs and the shear wave velocities within each of the 3 mm ROIs was recorded. The radiofrequency (RF) data used to produce the shear wave data online were stored on a hard drive for subsequent data acquisition using an offline version of the VTIQ software and RadiAnt DICOM Viewer (version 2021.2.2).

SWE images of the base of the tongue and suprahyoid neck were obtained in the transverse orientation at midline (anterior digastric and geniohyoid muscles) and left/right lateral infrahyoid neck in the longitudinal plane (sternohyoid and thyrohyoid muscles) for subsequent offline analyses. Up to three (minimum two) measurement ROIs were placed within each muscle belly, depending on limitations of muscle size and thickness. The sternothyroid was omitted from the SWE analysis because the muscle belly was thin and the smallest ROI markers overlapped with surrounding cartilaginous laryngeal framework. Shear wave velocities were acquired for ELMs at pre- and post-vocal load time points, unvoiced (baseline rest) and voiced (sustained steady state vocalization productions of /i/ (“eee” as in “neat”) at comfortable pitch and loudness.

See details of data acquisition for laryngoscopy, voice samples, the vocal load task, and self-perception ratings in the Supplemental Methods-Data Acquisition document.

See Figure 2 for study protocol schematic.

Figure 2.

Figure 2.

Open in a new tab

Study Protocol Schematic

Data Analysis

Shear Wave Elastography (ELM tension).

Stiffness measurements from the two-three ROIs for each of the three sets of trials for each muscle were obtained and averaged to represent one data point for each ELM, side, imaging plane, voicing (baseline unvoiced, voiced), and vocal load condition (pre-load, post-load) combination. The main effects of group and vocal load, as well as group × vocal load interactions on ELM tension, as measured by shear wave velocities, were determined with a 2 × 2 ANOVA and corrected with Bonferroni correction to reduce Type 1 error, with alpha set at 0.025 to account for multiple comparisons. We also calculated intraclass correlation coefficients (ICCs) across each set of three trials for each SWE measurement acquisition to determine intra-reliability across the three SWE acquisition trials within each side/plane/voicing/load combination.

Laryngoscopic Assessment (supraglottic compression severity).

Endolaryngeal outlet areas (in pixels) were obtained from the traced endolaryngeal boundaries during voiced segments (steady state). The endolaryngeal outlets were normalized to the average width (in pixels) of the endolarynx obtained during the same image taken during vocalization (Figure 3B). Mediolateral supraglottic compression severity was determined as (LO/Ŵ2) × 100, where LO is the endolaryngeal outlet area and W is the average width of the outlet (in pixels) (Figure 3B). The main effects of group and vocal load, as well as group × vocal load interactions using these metrics, were determined with a 2 × 2 ANOVA and corrected with Bonferroni correction to reduce Type 1 error.

Figure 3. Supraglottic Compression Schematic.

Figure 3.

Open in a new tab

(A) Endolaryngeal outlets for each image were traced manually using the boundaries between the true vocal folds, ventricular folds, interarytenoid mucosa, and petiole of the epiglottis. (B) The width of each endolaryngeal outlet (light blue spaces) perpendicular to every pixel along the length of the anterior-posterior endolarynx outlet (red line) was determined and averaged to obtain the mean width (dark blue line) for normalization. Mediolateral supraglottic compression severity was determined by (LO/Ŵ2) × 100, where LO is the endolaryngeal outlet area and W is the average width of the outlet (in pixels).

See details of data analysis for laryngoscopic assessment (mediolateral supraglottic compression severity), acoustic analysis of voice samples (cepstral peak prominence), and self-perception ratings of vocal effort and vocal tract discomfort in the Supplemental Methods-Data Analysis document.

Correlations Across Outcome Variables.

Relationships across suprahyoid (anterior digastric, geniohyoid) and infrahyoid (thyrohyoid, sternohyoid) ELM tension, supraglottic compression (laryngoscopy), acoustics (CPP), vocal effort (100 mm VAS), and vocal tract discomfort (VTDS) were compared using Pearson’s correlations.

RESULTS

Extrinsic Laryngeal Muscle Tension (shear wave).

ICCs for SWE measurements acquired in triplicates were highly reliable (p=0.0004). Both groups had significantly more tension during vocal production compared to baseline rest across all ELMs (p<0.0001), confirming the utility of the use of SWE for ELM tension to distinguish muscle activation from baseline rest to vocalization conditions (Figure 4A). The severity of ELM tension was similar between participants with pMTD and typical occupational voice users. There were no statistically significant group differences in any of the four ELMs (anterior digastric, geniohyoid, thyrohyoid, and sternohyoid) at baseline or during voicing at either the pre- or post-vocal load time points (p>0.05) (Figure 4B). There were no significant interactions between group and voicing (p=0.11), group and vocal load (p=0.44), or voicing and vocal load (p=0.45) for any of the ELMs.

Figure 4. Shear wave elastography velocity measures of extrinsic laryngeal muscle tension.

Figure 4.

Open in a new tab

(A) Effects of vocalization from the rest to voiced conditions, divided by pre- and post-vocal load time points. Both groups had significantly higher levels of ELM tension during the voiced conditions compared to baseline rest measurements at both time points across all ELMs. However, the level of change between rest and voiced conditions were similar between groups. The one exception was significantly more stiffness from baseline rest to vocalization in the left thyrohyoid in patients with pMTD compared to controls. This pattern could be representative of greater thyrohyoid space changes with voicing previously observed in patients with pMTD.51 However, due to the multiple statistical comparisons, and because this pattern only occurred on one side, it is likely that this lateralized pattern has more to do with individual variability than group significance. (B) Effect of vocal load at rest (left) and during voiced (right) conditions; each graph is divided by pre- and post-load time points. Shear wave velocities of extrinsic laryngeal muscle tension in the anterior digastric, geniohyoid, sternohyoid, and thyrohyoid were similar between occupational voice users with and without pMTD during rest and voicing conditions, at both pre- and post-vocal load time points.

Laryngoscopic Patterns of Supraglottic Hyperfunction.

Mediolateral supraglottic compression was significantly higher in the pMTD group than in the control group (p<0.0001). Vocal load did not have a significant effect on the severity of mediolateral supraglottic compression (p=0.48). There were also no interaction effects between group and vocal load on mediolateral supraglottic compression (p=0.69) (Figure 5).

Figure 5. Mediolateral Supraglottic Compression Severity.

Figure 5.

Open in a new tab

Occupational voice users with pMTD exhibited significantly more severe mediolateral supraglottic compression than typical voice users at both pre- and post-vocal load time points. (***p<.0001).

Acoustic Measures of Vocal Perturbation.

CPP was significantly lower in the pMTD group than in the control group at both pre-vocal load (p=0.01) and post-vocal load time points (p=0.0004). However, vocal load did not have a significant effect on CPP (p=0.95). No interaction effects between group and vocal load on CPP were observed (p=0.36). Although CPP was lower in the experimental group, both group means fell within the normative range of CPP values using similar methods previously used to acquire normative data in vocally healthy individuals.35 CPP values for healthy controls were closer to normophonic values while CPP values for patients with pMTD were closer to subclinical values.36 Twenty-one subjects out of 30 subjects in the pMTD group and 33 out of 36 subjects in the control group were above the 10.84 dB lower boundary of normophonic35 values (pre-vocal load: control - 15.13 ± 2.47 dB, pMTD - 13.31 ± 3.20 dB; post-vocal load: control - 15.63 ± 2.18 dB, pMTD - 12.87 ± 3.74 dB), similar to what is to be expected in patients with pMTD6 (Figure 6).

Figure 6. Cepstral Peak Prominence.

Figure 6.

Open in a new tab

Occupational voice users with pMTD exhibited significantly lower CPP values at pre-vocal load (*p<.05) and post-vocal load time points (***p<.0001) than typical occupational voice users. However, mean and standard deviation CPP values were within the normal range for both groups.

Self-Perception of Vocal Effort and Discomfort.

Participants with pMTD rated their vocal effort (p<0.0001) and vocal tract discomfort (p<0.0001) significantly higher than controls. Vocal load also had a significant effect on ratings of vocal effort (p<0.0001) and vocal tract discomfort (p<0.0001), suggesting that both groups felt that they were adequately vocally loaded. However, there were no interaction effects between group and vocal load time point for either vocal effort (p=0.70) or vocal tract discomfort (frequency, p=0.98; severity, p=0.94; total, p=0.96) (Figure 7).

Figure 7. Vocal Effort and Vocal Tract Discomfort.

Figure 7.

Open in a new tab

Occupational voice users with pMTD exhibited significantly higher vocal fold effort and vocal tract discomfort at pre- and post-vocal load time points (***p<.0001). Both groups reported significantly more vocal effort and vocal tract discomfort at post-vocal load compared to pre-vocal load (**p<.01).

See Supplementary Table 1 with descriptive statistics for each of the outcome variables.

Correlations Across Variables.

There was a significant positive correlation between infrahyoid tension and CPP, but not suprahyoid muscles, in both the pMTD group and control group at baseline. After vocal load, the pMTD group had significant positive correlations in both suprahyoid and infrahyoid groups of ELMs and CPP, while the control group only had significant positive correlations between sternohyoid and CPP. There were no significant relationships between ELM tension and vocal effort in either pMTD or control group at the pre-vocal load time point. However, the pMTD group had a significant negative correlation between suprahyoid and infrahyoid ELM tension and effort at the post-vocal load time point. There were also significant negative correlations between infrahyoid tension and vocal effort in the control group after the vocal load. There were no significant relationships between ELM tension and vocal tract discomfort, except for negative correlations found between the infrahyoid muscles and vocal tract discomfort in the pMTD group at the post-vocal load time point. There were no significant relationships between CPP and vocal effort and CPP and vocal tract discomfort for any of the groups and time points except for a negative correlation between CPP and vocal effort in the pMTD group at the post-load time point. Mediolateral supraglottic compression did not have any significant correlations to ELM tension or other standard voice instrumentation metrics. There were significant positive correlations between vocal effort and vocal tract discomfort across both groups and timepoints (see Supplemental Tables 2–5 for Pearson’s correlations and p-values).

DISCUSSION

Results demonstrate the utility of SWE for quantification of muscle tension in the ELMs. SWE measurements were highly reliable and were able to distinguish muscle activation in the ELMs from baseline rest to vocalization in both groups. The results of the study also showed that the two groups had similar SWE patterns, in two distinct ways. First, the SWE measures of ELM tension at baseline, during vocal production, and pre- versus post-vocal load were comparable between the groups. Second, there were no significant group differences in muscle tension measured using SWE from unvoiced to voiced productions. Taken together, these results demonstrate that voice users with pMTD do not have more quantifiable ELM tension than typical occupational voice users. The lack of differences in quantitative measures between the groups corresponds to previously noted inconclusive findings with other ELM assessment methods (e.g., electromyography,5,37–40 palpation41).

These findings were further corroborated by the lack of relationships between ELM tension and vocal effort and vocal tract discomfort in both groups at the pre-vocal load time point. These data suggest muscle tension in the ELMs and patient sensory perceptions of increased vocal effort and vocal tract discomfort might be distinct entities. In fact, there were significant negative correlations between suprahyoid and infrahyoid ELM tension and vocal effort/discomfort after the vocal load task in the pMTD group, which further refutes conventional theories of the role of ELM tension in patients with pMTD, even after heavy vocal demands.

There are three possible interpretations of muscle pattern similarities between groups. The first is that the patients with pMTD were not severe enough in their clinical presentations for SWE to pick up subtleties in levels of ELM tension. However, this cohort was representative of clinical populations typically seen in the voice clinic and were all patients who were subjectively diagnosed with muscle tension. As such, quantification of muscle tension with SWE should parallel these subjective assessments. Considering SWE was able to successfully quantify increased ELM tension with vocalization from baseline conditions in this study, as well as limb muscle tension in various previous limb muscle studies, then it should also theoretically be able to distinguish differences in level of muscle tension in the neck between groups if they do exist.

The second possibility is that muscle tension occurred in both groups of occupational voice users and may instead reflect chronic high vocal demands instead of a clinical indication of pMTD pathophysiology. Stated differently, what clinicians may observe is a consequence of occupational voice use and not vocal dysfunction. The presence of these muscle patterns may be similar to tighter or more active limb muscles more prevalent in athletes than sedentary individuals. To elucidate this theory, future studies should compare SWE methods in occupational and non-occupational voice users with and without pMTD.

The third possibility is that ascribing ELM tension to pMTD is the consequence of the anchoring effect, a cognitive bias that results in heavy reliance on initial information received.42 It may be that increased tension in the ELMs do not actually exist in patients with pMTD or that these patterns occur variably across individuals with pMTD and typical voice users (for example, as a natural compensation29, which is supported by the positive correlations between ELM tension and CPP values in the present study). Instead, we may be relying on initial, untested theories and interpreting and adapting newer information from this erroneous reference point. The literature from which we have anchored often omit group comparisons and depend on descriptive clinical observations and expert option.2,28,43

Results from the few studies that have included more objective metrics tell a story similar to the findings of the present study. For example, Hirano (1969) found similar levels of sternohyoid tension between “optimal” and “hyperfunctional” vocal productions using hook-wire electromyography.44 Furthermore, although level of sternothyroid, sternohyoid, and thyrohyoid electromyographic muscle activation have been observed with pitch changes,45–47 pitch deviations (too high or too low) are not central characteristics in patients with pMTD (as indicated by previous work6,7 and fundamental frequencies within normal limits found in the present study; c.f., Table 1).

Despite the similarities in ELM tension patterns between the groups, voice users with pMTD experienced significantly higher self-perceptual ratings of vocal effort and vocal tract discomfort at both the pre- and post-vocal load time points. Measures of vocal effort and vocal tract discomfort also had the strongest correlations. Furthermore, participants with pMTD had significantly higher mean mediolateral supraglottic compression severity on laryngoscopy and lower (although relatively normal) CPP values. There are several interpretations of these findings that all suggest pMTD may be more sensory in nature than originally considered. Greater sensory perceptions of vocal effort and vocal tract discomfort, despite similar ELM tension patterns between groups, suggest that patients with pMTD have different sensory experiences from those of typical voice users.

The relatively normal acoustic CPP values in both groups may further support the idea that pMTD may have more to do with the feel of voicing (i.e., sensation) than the sound. The lack of significant relationships between CPP values and measures of vocal effort and vocal tract discomfort at the pre-vocal load time point further support this theory. Greater mediolateral supraglottic compression in patients with pMTD may point to a hyper-responsive airway protection mechanism in some individuals with pMTD, previously observed in disorders of laryngeal hypersensitivity (e.g., irritable larynx syndrome).48–50 Interestingly, the lack of correlations between supraglottic compression and other outcome variables suggests these supraglottic compression patterns may co-occur with laryngeal hypersensitivity but may be distinct from sensory experiences of vocal effort and discomfort in patients with pMTD. However, further studies are needed to elucidate these hypotheses.

Several limitations should be noted in this study. The first is that muscle tension was only measured at baseline, during steady state vowel productions, and at modal pitch. As such, muscle activation prior to, and immediately at onset, was not observed. The effects of pitch changes on ELM tension using SWE between the two groups were also not observed. The second is that degree of muscle tension was measured during vowel productions, but not speech production. Muscle tension levels may differ between groups during speech tasks and sustained vowels on modal pitch and intensity may not be enough to capture ELM tension pathophysiology. Future shear wave studies that look at muscle tension at vocal onset, during speech production, and at different pitches are needed.

CONCLUSION

Shear wave elastrography was able to distinguish muscle activation with vocalization compared to at-rest conditions, but did not distinguish ELM tension levels between groups with and without pMTD. Occupational voice users with pMTD reported significantly greater levels of vocal effort and vocal tract discomfort, and on average, exhibited significantly more mediolateral supraglottic compression and lower CPP values than typical occupational voice users. However, these clinical presentations were not supported by the similar quantitative patterns of ELM tension on SWE between the two groups. Correlations between increased ELM tension and increased sensory experiences of vocal effort and vocal tract discomfort were also poor. Considering these findings in the context of previous work to assess the ELMs (e.g., electromyography and palpation), the results of this study raise the question of the validity of ELM tension as a clinical indicator of pMTD, although further studies are need to support or refute this claim. These findings also do not support the theory that common symptoms of vocal effort, vocal fatigue, vocal tract discomfort, odynophonia that patients with pMTD experience are the result of ELM tension.

Supplementary Material

Supinfo1

Supplementary Methods. Data Acquisition

Supinfo2

Supplementary Methods. Data Analysis

t2

Supplementary Table 2. Pearson Correlations for Experimental Group (pMTD) Pre-Vocal Load.

t1

Supplemental Table 1. Descriptive Statistics for shear wave measures, vocal function outcomes, and self-perceptual measures across pre-post vocal load time points.

t3

Supplementary Table 3. Pearson Correlations for Control Group Pre-Vocal Load

t4

Supplementary Table 4. Pearson Correlations for Experimental Group (pMTD) Post-Vocal Load.

t5

Supplementary Table 5. Pearson Correlations for Control Group Post-Vocal Load

Acknowledgements:

Special thanks to Lesley Childs, Laura Toles, and Amy Harris for their help with subject recruitment; Crystal Bruce RDMS RVT RDCS, Simon Giang RDMS RVT, and the rest of the UT Southwestern CACTUS Lab for their help with shear wave ultrasound acquisition; Katie Bosler, Jenny Maique, and Asha Varghese for assistance with the study protocol runs; Jasper Han for creating the customized ImageJ macro for quantitative supraglottic configuration analysis, the Spring 2023 COMD 6221 Voice Disorders class for their help with data analysis and interpretation, Youri Maryn, PhD at Phonanium CommV for providing the customized Praat script for acoustic analysis, and Ronald Adler and Iman Khodarahmi at NYU Langone Health for their help with initial conceptualization and study design.

Funding:

This work was supported by NIDCD grant R21DC019207. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Deafness and Other Communication Disorders or the National Institutes of Health.

Footnotes

Level of evidence: 2

Conflict of Interest: Author D.T.F. -- Research Agreements, GE Healthcare, Philips Healthcare, and Siemens Healthineers; Advisory Board, GE Healthcare and Philips Healthcare. The other authors have no conflicts of interests to disclose.

Data Availability:

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

References

  • 1.Desjardins M, Apfelbach C, Rubino M, Verdolini AK. Integrative Review and Framework of Suggested Mechanisms in Primary Muscle Tension Dysphonia. Journal of Speech, Language, and Hearing Research. doi: 10.1044/2022_JSLHR-21-00575 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Morrison MD, Rammage LA. Muscle Misuse Voice Disorders: Description and Classification. Acta Oto-Laryngologica. 1993;113(3):428–434. doi: 10.3109/00016489309135839 [DOI] [PubMed] [Google Scholar]
  • 3.Tierney WS, Xiao R, Milstein CF. Characterization of Functional Dysphonia: Pre- and Post-Treatment Findings. The Laryngoscope. n/a(n/a). doi: 10.1002/lary.29358 [DOI] [PubMed] [Google Scholar]
  • 4.Khoddami SM, Ansari NN, Jalaie S. Review on Laryngeal Palpation Methods in Muscle Tension Dysphonia: Validity and Reliability Issues. Journal of Voice. 2015;29(4):459–468. doi: 10.1016/j.jvoice.2014.09.023 [DOI] [PubMed] [Google Scholar]
  • 5.Van Houtte E, Claeys S, D’haeseleer E, Wuyts F, Van Lierde K. An Examination of Surface EMG for the Assessment of Muscle Tension Dysphonia. Journal of Voice. 2013;27(2):177–186. doi: 10.1016/j.jvoice.2011.06.006 [DOI] [PubMed] [Google Scholar]
  • 6.Shembel AC, Lee J, Sacher JR, Johnson AM. Characterization of Primary Muscle Tension Dysphonia Using Acoustic and Aerodynamic Voice Metrics. Journal of Voice. Published online July 17, 2021. doi: 10.1016/j.jvoice.2021.05.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Gillespie AI, Gartner-Schmidt J. Immediate Effect of Stimulability Assessment on Acoustic, Aerodynamic, and Patient-Perceptual Measures of Voice. Journal of Voice. 2016;30(4):507.e9–507.e14. doi: 10.1016/j.jvoice.2015.06.004 [DOI] [PubMed] [Google Scholar]
  • 8.Davis LC, Baumer TG, Bey MJ, van Holsbeeck M. Clinical utilization of shear wave elastography in the musculoskeletal system. Ultrasonography. 2019;38(1):2–12. doi: 10.14366/usg.18039 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Creze M, Nordez A, Soubeyrand M, Rocher L, Maître X, Bellin MF. Shear wave sonoelastography of skeletal muscle: basic principles, biomechanical concepts, clinical applications, and future perspectives. Skeletal Radiol. 2018;47(4):457–471. doi: 10.1007/s00256-017-2843-y [DOI] [PubMed] [Google Scholar]
  • 10.Akagi R, Sato S, Yoshihara K, Ishimatsu H, Ema R. Sex difference in fatigability of knee extensor muscles during sustained low-level contractions. Sci Rep. 2019;9(1):16718. doi: 10.1038/s41598-019-53375-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Bouillard K, Jubeau M, Nordez A, Hug F. Effect of vastus lateralis fatigue on load sharing between quadriceps femoris muscles during isometric knee extensions. Journal of Neurophysiology. 2013;111(4):768–776. doi: 10.1152/jn.00595.2013 [DOI] [PubMed] [Google Scholar]
  • 12.Brandenburg JE, Eby SF, Song P, et al. Ultrasound Elastography: The New Frontier in Direct Measurement of Muscle Stiffness. Arch Phys Med Rehabil. 2014;95(11):2207–2219. doi: 10.1016/j.apmr.2014.07.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Flatres A, Aarab Y, Nougaret S, et al. Real-time shear wave ultrasound elastography: a new tool for the evaluation of diaphragm and limb muscle stiffness in critically ill patients. Critical Care. 2020;24(1):34. doi: 10.1186/s13054-020-2745-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Gennisson JL, Deffieux T, Macé E, Montaldo G, Fink M, Tanter M. Viscoelastic and Anisotropic Mechanical Properties of in vivo Muscle Tissue Assessed by Supersonic Shear Imaging. Ultrasound in Medicine & Biology. 2010;36(5):789–801. doi: 10.1016/j.ultrasmedbio.2010.02.013 [DOI] [PubMed] [Google Scholar]
  • 15.Hug F, Tucker K, Gennisson JL, Tanter M, Nordez A. Elastography for Muscle Biomechanics: Toward the Estimation of Individual Muscle Force. Exercise and Sport Sciences Reviews. 2015;43(3):125. doi: 10.1249/JES.0000000000000049 [DOI] [PubMed] [Google Scholar]
  • 16.Liu J, Qian Z, Wang K, et al. Non-invasive Quantitative Assessment of Muscle Force Based on Ultrasonic Shear Wave Elastography. Ultrasound in Medicine and Biology. 2019;45(2):440–451. doi: 10.1016/j.ultrasmedbio.2018.07.009 [DOI] [PubMed] [Google Scholar]
  • 17.Siracusa J, Charlot K, Malgoyre A, et al. Resting Muscle Shear Modulus Measured With Ultrasound Shear-Wave Elastography as an Alternative Tool to Assess Muscle Fatigue in Humans. Front Physiol. 2019;10. doi: 10.3389/fphys.2019.00626 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Lacourpaille L, Hug F, Guével A, et al. Non-invasive assessment of muscle stiffness in patients with duchenne muscular dystrophy. Muscle & Nerve. 2015;51(2):284–286. doi: 10.1002/mus.24445 [DOI] [PubMed] [Google Scholar]
  • 19.Lacourpaille L, Nordez A, Hug F, Couturier A, Dibie C, Guilhem G. Time-course effect of exercise-induced muscle damage on localized muscle mechanical properties assessed using elastography. Acta Physiologica. 2014;211(1):135–146. doi: 10.1111/apha.12272 [DOI] [PubMed] [Google Scholar]
  • 20.Hug F, Ouellette A, Vicenzino B, Hodges PW, Tucker K. Deloading tape reduces muscle stress at rest and during contraction. Med Sci Sports Exerc. 2014;46(12):2317–2325. doi: 10.1249/mss.0000000000000363 [DOI] [PubMed] [Google Scholar]
  • 21.Eriksson Crommert M, Lacourpaille L, Heales LJ, Tucker K, Hug F. Massage induces an immediate, albeit short-term, reduction in muscle stiffness. Scandinavian Journal of Medicine & Science in Sports. 2015;25(5):e490–e496. doi: 10.1111/sms.12341 [DOI] [PubMed] [Google Scholar]
  • 22.Lopes LW, Cabral GF, Figueiredo de Almeida AA. Vocal Tract Discomfort Symptoms in Patients With Different Voice Disorders. Journal of Voice. 2015;29(3):317–323. doi: 10.1016/j.jvoice.2014.07.013 [DOI] [PubMed] [Google Scholar]
  • 23.Galletti B, Sireci F, Mollica R, et al. Vocal Tract Discomfort Scale (VTDS) and Voice Symptom Scale (VoiSS) in the Early Identification of Italian Teachers with Voice Disorders. Int Arch Otorhinolaryngol. 2020;24(3):e323–e329. doi: 10.1055/s-0039-1700586 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Smolander S, Huttunen K. Voice problems experienced by Finnish comprehensive school teachers and realization of occupational health care. Logopedics Phoniatrics Vocology. 2006;31(4):166–171. doi: 10.1080/14015430600576097 [DOI] [PubMed] [Google Scholar]
  • 25.Sliwinska-Kowalska M, Niebudek-Bogusz E, Fiszer M, et al. The Prevalence and Risk Factors for Occupational Voice Disorders in Teachers. Folia Phoniatrica et Logopaedica. 2006;58(2):85–101. doi: 10.1159/000089610 [DOI] [PubMed] [Google Scholar]
  • 26.Boucher VJ. Acoustic Correlates of Fatigue in Laryngeal Muscles: Findings for a Criterion-Based Prevention of Acquired Voice Pathologies. J Speech Lang Hear Res. 2008;51(5):1161–1170. doi: 10.1044/1092-4388(2008/07-0005) [DOI] [PubMed] [Google Scholar]
  • 27.D’haeseleer E, Behlau M, Bruneel L, et al. Factors Involved in Vocal Fatigue: A Pilot Study. Folia Phoniatr Logop. 2016;68(3):112–118. doi: 10.1159/000452127 [DOI] [PubMed] [Google Scholar]
  • 28.Verdolini K, Rosen C, Branski RC. Classification manual for voice disorders-I, Special Interest Division 3, Voice and Voice disorders. American Speech-Language Hearing Division. Published online 2005. [Google Scholar]
  • 29.McDowell S, Morrison R, Mau T, Shembel AC. Clinical characteristics and effects of vocal demands in occupational voice users with and without primary muscle tension dysphonia. Journal of Voice. Published online In Proof 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Nanjundeswaran C, Shembel AC. Laying the Groundwork to Study the Heterogeneous Nature of Vocal Fatigue. Journal of Voice. Published online August 6, 2022. doi: 10.1016/j.jvoice.2022.07.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Smeltzer JC, Chiou SH, Shembel AC. Interoception, Voice Symptom Reporting, and Voice Disorders. Journal of Voice. Published online April 1, 2023. doi: 10.1016/j.jvoice.2023.03.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Rosen CA, Lee AS, Osborne J, Zullo T, Murry T. Development and Validation of the Voice Handicap Index-10. The Laryngoscope. 2004;114(9):1549–1556. [DOI] [PubMed] [Google Scholar]
  • 33.Arffa RE, Krishna P, Gartner-Schmidt J, Rosen CA. Normative Values for the Voice Handicap Index-10. Journal of Voice. 2012;26(4):462–465. doi: 10.1016/j.jvoice.2011.04.006 [DOI] [PubMed] [Google Scholar]
  • 34.Nanjundeswaran C, Jacobson BH, Gartner-Schmidt J, Verdolini Abbott K. Vocal Fatigue Index (VFI): Development and Validation. J Voice. 2015;29(4):433–440. doi: 10.1016/j.jvoice.2014.09.012 [DOI] [PubMed] [Google Scholar]
  • 35.Murton O, Hillman R, Mehta D. Cepstral Peak Prominence Values for Clinical Voice Evaluation. Am J Speech Lang Pathol. Published online July 13, 2020:1–12. doi: 10.1044/2020_AJSLP-20-00001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Batthyany C, Maryn Y, Trauwaen I, et al. A Case of Specificity: How Does the Acoustic Voice Quality Index Perform in Normophonic Subjects? Applied Sciences. 2019;9(12):2527. doi: 10.3390/app9122527 [DOI] [Google Scholar]
  • 37.Hocevar-Boltezar I, Janko M, Zargi M. Role of surface EMG in diagnostics and treatment of muscle tension dysphonia. Acta Otolaryngologica. 1998;118:739–743. [DOI] [PubMed] [Google Scholar]
  • 38.Redenbaugh Margaret A, Reich Alan R Surface EMG and Related Measures in Normal and Vocally Hyperfunctional Speakers. Journal of Speech and Hearing Disorders. 1989;54(1):68–73. doi: 10.1044/jshd.5401.68 [DOI] [PubMed] [Google Scholar]
  • 39.Stemple JC, Weiler E, Whitehead W, Komray R. Electromyographic biofeedback training with patients exhibiting a hyperfunctional voice disorder. The Laryngoscope. 1980;90(3):471–476. [DOI] [PubMed] [Google Scholar]
  • 40.Stepp ACE, Hillman RE, Heaton JT. Use of Neck Strap Muscle Intermuscular Coherence as an Indicator of Vocal Hyperfunction. IEEE Trans Neural Syst Rehabil Eng. 2010;18(3):329–335. doi: 10.1109/TNSRE.2009.2039605 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Spencer ML. Muscle Tension Dysphonia: A Rationale for Symptomatic Subtypes, Expedited Treatment, and Increased Therapy Compliance. Perspect Voice Voice Dis. 2015;25(1):5–15. doi: 10.1044/vvd25.1.5 [DOI] [Google Scholar]
  • 42.Furnham A, Boo HC. A literature review of the anchoring effect. The Journal of Socio-Economics. 2011;40(1):35–42. doi: 10.1016/j.socec.2010.10.008 [DOI] [Google Scholar]
  • 43.Koufman JA, Blalock PD. Functional voice disorders. Otolaryngol Clin North Am. 1991;24(5):1059–1073. [PubMed] [Google Scholar]
  • 44.Hirano M, Koike Y, Joyner J. Style of Phonation: An Electromyographic Investigation of Some Laryngeal Muscles. Arch Otolaryngol. 1969;89(6):902–907. doi: 10.1001/archotol.1969.00770020904020 [DOI] [PubMed] [Google Scholar]
  • 45.Vilkman E, Sonninen A, Hurme P, Körkkö P. External laryngeal frame function in voice production revisited: A review. Journal of Voice. 1996;10(1):78–92. doi: 10.1016/S0892-1997(96)80021-X [DOI] [PubMed] [Google Scholar]
  • 46.Sokolowsky RR. EFFECT OF THE EXTRINSIC LARYNGEAL MUSCLES ON VOICE PRODUCTION. Archives of Otolaryngology - Head and Neck Surgery. 1943;38(4):355–364. doi: 10.1001/archotol.1943.00670040369008 [DOI] [Google Scholar]
  • 47.Andersen KF, Sonninen A. The Function of the Extrinsic Laryngeal Muscles at Different Pitch. Acta Oto-Laryngologica. 1960;51(1–2):89–93. doi: 10.3109/00016486009124468 [DOI] [PubMed] [Google Scholar]
  • 48.Morrison M, Rammage L. The irritable larynx syndrome as a central sensitivity syndrome. Canadian Journal of Speech-Language Pathology and Audiology. 2010;34(4). [Google Scholar]
  • 49.Vertigan AE, Bone SL, Gibson PG. Laryngeal sensory dysfunction in laryngeal hypersensitivity syndrome. Respirology. 2013;18(6):948–956. doi: 10.1111/resp.12103 [DOI] [PubMed] [Google Scholar]
  • 50.Morrison M, Rammage L, Emami AJ. The irritable larynx syndrome. Journal of Voice. 1999;13(3):447–455. [DOI] [PubMed] [Google Scholar]
  • 51.Lowell SY, Kelley RT, Colton RH, Smith PB, Portnoy JE. Position of the hyoid and larynx in people with muscle tension dysphonia. The Laryngoscope. 2012;122(2):370–377. doi: 10.1002/lary.22482 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supinfo1

Supplementary Methods. Data Acquisition

NIHMS1909131-supplement-Supinfo1.docx (15KB, docx)
Supinfo2

Supplementary Methods. Data Analysis

NIHMS1909131-supplement-Supinfo2.docx (15.8KB, docx)
t2

Supplementary Table 2. Pearson Correlations for Experimental Group (pMTD) Pre-Vocal Load.

NIHMS1909131-supplement-t2.docx (19.3KB, docx)
t1

Supplemental Table 1. Descriptive Statistics for shear wave measures, vocal function outcomes, and self-perceptual measures across pre-post vocal load time points.

NIHMS1909131-supplement-t1.pdf (92.8KB, pdf)
t3

Supplementary Table 3. Pearson Correlations for Control Group Pre-Vocal Load

NIHMS1909131-supplement-t3.docx (19KB, docx)
t4

Supplementary Table 4. Pearson Correlations for Experimental Group (pMTD) Post-Vocal Load.

NIHMS1909131-supplement-t4.docx (19.1KB, docx)
t5

Supplementary Table 5. Pearson Correlations for Control Group Post-Vocal Load

NIHMS1909131-supplement-t5.docx (19.1KB, docx)

Data Availability Statement

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

RESOURCES