Development of Imaging Biomarkers for Voice After Pediatric Airway Reconstruction

0 subjects completed all study visits. Due to noncompliance and recruitment difficulties study was terminated.

Objective: The purpose of this proposal is to use High Speed Videoendoscopy (HSV) to develop and test an imaging biomarker to support diagnosis and predict vibratory outcome after airway reconstruction. Our protocol will involve eliciting and recording the following five behaviors: (1) hard throat clear; (2) cough; (3) laugh; (4) short multiple phonations; and (5) sustained phonation. These HSV recordings will be used to develop the biomarkers. Study Design: A prospective cohort study of 60 patients, 36 who will undergo airway reconstruction and 24 age matched controls. Setting: Center for Pediatric Voice Disorders, Department of Otolaryngology, Cincinnati Children's Hospital Medical Center Methods: Subjects will undergo a full voice evaluation, including an HSV evaluation protocol, at the main study site prior to airway reconstruction, 6 months post and 1 year post surgery. Voice evaluation will include aerodynamic, acoustic, and perceptual voice assessments, a pediatric voice handicap index questionnaire, videostroboscopy, and HSV. The airway evaluation will include airway sizing and calibration.

Fifty percent of the children who undergo airway reconstruction for subglottic stenosis (narrowing of the airway below the vocal folds) develop a voice disorder. Voice disorders lead to psychosocial problems that affect many of these children over their lifetime, specifically their career choices and their long-term quality of life. Our current understanding of voice problems following surgical airway reconstruction is limited. There is a lack of research to understand the alterations of the phonatory mechanisms due to the underlying initial condition (stenosis) and the surgical airway reconstruction performed. The purpose of this proposal is to develop and test imaging biomarkers to support diagnosis and surgical decision-making leading to improved voice outcome. The overarching goal of this work is to establish a foundation upon which surgical practice could be enhanced to improve voice outcome while still achieving optimal airway results. Specifically, the project will use: a prospective pre- and post-surgery cohort design to develop an imaging biomarker of "vibratory potential" for predicting the vibratory outcome after airway reconstruction. Quantitative image processing methodology will be used for the development and validation of the new imaging biomarkers, based on high-speed videoendoscopy. This research project will address a substantial gap of knowledge in pediatric voice and airway surgery, and has the potential to significantly improve the quality of life of these patients who have some of the most severe and difficult to manage voice disorders. The results could be translated to adult patient populations and other voice disorders (e.g. development of disease-specific imaging biomarkers).

These participants have significant airway abnormalities that require invasive surgery, such as Laryngotracheoplasty, to rectify, and whose voice quality may suffer as a result of the surgery. The goal of this study is to improve voice outcomes of these patients, and their clinical data will be collected.

These participants have normal airways and voice who will undergo a microlaryngoscopy and voice evaluation, the data from which will be compared to study patients.

Standard clinical procedure, used to perform airway sizing and vocal fold calibration. Controls receive one as a research procedure, while study patients receive one as part of standard of care.

Standard clinical procedure that includes aerodynamic, acoustic, and perceptual voice assessments, a pediatric voice handicap index questionnaire, videostroboscopy, and high-speed videoendoscopy. Controls will receive one as a research procedure, while study patients will receive three, one before their airway reconstruction, and two more in the year afterward, as part of standard of care.

This includes many types of invasive surgical procedures such as Laryngotracheoplasty to repair a patient's compromised airway. Only study patients will be receiving one, and it will be as part of standard of care unrelated to this study, but the data from surgery and outcomes being pertinent to the study.

Vibratory source will be rated by having HSV imaging samples subjected to an automatic temporal segmentation, which will extract the HSV segment of interest. The extracted temporal segment (i.e. token) is subjected to an automatic image segmentation based on the oscillating edges of the anatomic structures using kymographic paired active contours. The image segmentation will be aided by an automatic spectral oscillation analysis similar to previously published methods. Three raters will be asked to interpret the image analysis results, which will be overlaid on the HSV image. To identify the patient's vibratory source, among the segmented vibratory areas, the raters will identify which vibratory structure(s)--glottis, false vocal folds, aryepiglottic folds, arytenoid against epiglottis, other supraglottal structures, or any combination of those--could be phonatory/which anatomic structure produced the vibration.

Tissue health will be rated using a specialized Matlab graphic user interface (GUI). The rating protocol for preoperative and postoperative measurement will assess vibratory features on an ordinal scale, as follows: Vocal-fold vibration will be rated as: vibration with contact; vibration, no contact; no vibration, with contact; no vibration, no contact; can't identify vocal folds; or can't rate. The raters will then be asked to grade their overall rating confidence in their rating. Confidence will be graded on a 5-point scale as follows: 5-highly confident; 4-confident; 3-mostly confident; 2-somewhat confident; and 1- not confident or unable to rate.

Tissue health will be rated using a specialized Matlab graphic user interface (GUI). The rating protocol for preoperative and postoperative measurement will assess vibratory features on an ordinal scale, as follows:Scarring will be rated as: no scar; mild-left; mild-right; mild-bilateral; moderate-left; moderate- right; moderate- bilateral; severe-left; severe-right; severe- bilateral; or can't rate. For the visual rating of scarring, we will use the following definitions: mild scarring - present scar that has no effect or mildly reduces the mucosal wave; moderate scarring - reduced mucosal wave; severe scarring - mucosal wave not present. The raters will then be asked to grade their overall rating confidence in their rating. Confidence will be graded on a 5-point scale as follows: 5-highly confident; 4-confident; 3-mostly confident; 2-somewhat confident; and 1- not confident or unable to rate.

The computation of the adduction and abduction will be based on automatic image segmentation using temporally-paired active contours followed by extraction of a glottal area width (GAW) signal. This measure will be calibrated to the units of mm based on the segmented airway size relative to the size of the calibration surgical instrument recorded on the intraoperative image. Maximum adduction will be computed from the GAW signal.

The computation of the adduction and abduction will be based on automatic image segmentation using temporally-paired active contours followed by extraction of a glottal area width (GAW) signal. This measure will be calibrated to the units of mm based on the segmented airway size relative to the size of the calibration surgical instrument recorded on the intraoperative image. Maximum abduction will be computed from the GAW signal.

Three certified speech-language pathologists with extensive experience in the assessment and treatment of pediatric voice disorders will perform all CAPE-V ratings. Each sample will contain the six CAPE-V sentences or modified sentences based on the ability of the child. Rating sheets will list only the participant's study number, age, and gender. Judges will independently listen to the samples (a maximum of three times per sample) in a quiet environment using headphones (the headphones will not have a separate volume control function. Averages of the three judges will be performed to obtain the Overall Severity rating. Twenty-five percent of the voice samples will be randomly selected to estimate intra-rater reliability. Inter-rater reliability will also be evaluated. If CAPE-V Overall Severity ratings differ greater than 20%, samples will be set aside and evaluated in a consensus of all three judges. This consensus will establish the final Overall Severity rating.

Questionnaire designed to assess the impact on quality of life that the patient's voice problems pose. This is administered during each voice evaluation to track changes. Voice handicapping/voice-related quality of life will be evaluated using the Pediatric Voice Handicap Index (pVHI). The pVHI is a handicapping index that was modified from the original Adult Voice Handicap Index into a parental proxy form. The pVHI is a self-rating of voice handicapping based on a summed score (0-92). This system has three domains: 1) functional, 2) physical, and 3) emotional.

Study Patients Inclusion Criteria: Ages of 3-30 years, No prior airway surgery Able to undergo the evaluation protocol Exclusion Criteria: 1.) Patients who have undergone any prior airway reconstruction and/or have a disability that precludes them from completing the evaluation Control Patients Inclusion Criteria: Age of 3-30 years No prior history of intubation injury or airway/laryngeal surgery Normal hearing Normal voice assessment performed by a certified speech-language pathologist 5.) Undergoing a non-airway surgical procedure such as adenoidectomy, tonsillectomy, ear tubes, and neck mass removal. Exclusion Criteria: Has had airway or laryngeal surgery History of a voice or hearing disorder History of neonatal intubation over 14 days Have a disability that precludes them from completing the evaluation