An introduction to machine learning and generative artificial intelligence for otolaryngologists-head and neck surgeons: a narrative review.

PURPOSE: Despite the robust expansion of research surrounding artificial intelligence (AI) and machine learning (ML) and their applications to medicine, these methodologies often remain opaque and inaccessible to many otolaryngologists. Especially, with the increasing ubiquity of large-language models (LLMs), such as ChatGPT and their potential implementation in clinical practice, clinicians may benefit from a baseline understanding of some aspects of AI. In this narrative review, we seek to clarify underlying concepts, illustrate applications to otolaryngology, and highlight future directions and limitations of these tools. METHODS: Recent literature regarding AI principles and otolaryngologic applications of ML and LLMs was reviewed via search in PubMed and Google Scholar. RESULTS: Significant recent strides have been made in otolaryngology research utilizing AI and ML, across all subspecialties, including neurotology, head and neck oncology, laryngology, rhinology, and sleep surgery. Potential applications suggested by recent publications include screening and diagnosis, predictive tools, clinical decision support, and clinical workflow improvement via LLMs. Ongoing concerns regarding AI in medicine include ethical concerns around bias and data sharing, as well as the "black box" problem and limitations in explainability. CONCLUSIONS: Potential implementations of AI in otolaryngology are rapidly expanding. While implementation in clinical practice remains theoretical for most of these tools, their potential power to influence the practice of otolaryngology is substantial.

Sociodemographic reporting in videomics research: a review of practices in otolaryngology - head and neck surgery.

OBJECTIVE: To assess reporting practices of sociodemographic data in Upper Aerodigestive Tract (UAT) videomics research in Otolaryngology-Head and Neck Surgery (OHNS). STUDY DESIGN: Narrative review. METHODS: Four online research databases were searched for peer-reviewed articles on videomics and UAT endoscopy in OHNS, published since January 1, 2017. Title and abstract search, followed by a full-text screening was performed. Dataset audit criteria were determined by the MINIMAR reporting standards for patient demographic characteristics, in addition to gender and author affiliations. RESULTS: Of the 57 studies that were included, 37% reported any sociodemographic information on their dataset. Among these studies, all reported age, most reported sex (86%), two (10%) reported race, and one (5%) reported ethnicity and socioeconomic status. No studies reported gender. Most studies (84%) included at least one female author, and more than half of the studies (53%) had female first/senior authors, with no significant differences in the rate of sociodemographic reporting in studies with and without female authors (any female author: p = 0.2664; first/senior female author: p > 0.9999). Most studies based in the US reported at least one sociodemographic variable (79%), compared to those in Europe (24%) and in Asia (20%) (p = 0.0012). The rates of sociodemographic reporting in journals of different categories were as follows: clinical OHNS: 44%, clinical non-OHNS: 40%, technical: 42%, interdisciplinary: 10%. CONCLUSIONS: There is prevalent underreporting of sociodemographic information in OHNS videomics research utilizing UAT endoscopy. Routine reporting of sociodemographic information should be implemented for AI-based research to help minimize algorithmic biases that have been previously demonstrated.

A deep learning pipeline for automated classification of vocal fold polyps in flexible laryngoscopy.

PURPOSE: To develop and validate a deep learning model for distinguishing healthy vocal folds (HVF) and vocal fold polyps (VFP) on laryngoscopy videos, while demonstrating the ability of a previously developed informative frame classifier in facilitating deep learning development. METHODS: Following retrospective extraction of image frames from 52 HVF and 77 unilateral VFP videos, two researchers manually labeled each frame as informative or uninformative. A previously developed informative frame classifier was used to extract informative frames from the same video set. Both sets of videos were independently divided into training (60%), validation (20%), and test (20%) by patient. Machine-labeled frames were independently verified by two researchers to assess the precision of the informative frame classifier. Two models, pre-trained on ResNet18, were trained to classify frames as containing HVF or VFP. The accuracy of the polyp classifier trained on machine-labeled frames was compared to that of the classifier trained on human-labeled frames. The performance was measured by accuracy and area under the receiver operating characteristic curve (AUROC). RESULTS: When evaluated on a hold-out test set, the polyp classifier trained on machine-labeled frames achieved an accuracy of 85% and AUROC of 0.84, whereas the classifier trained on human-labeled frames achieved an accuracy of 69% and AUROC of 0.66. CONCLUSION: An accurate deep learning classifier for vocal fold polyp identification was developed and validated with the assistance of a peer-reviewed informative frame classifier for dataset assembly. The classifier trained on machine-labeled frames demonstrates improved performance compared to the classifier trained on human-labeled frames.

Usefulness, acceptation and feasibility of electronic medical history tool in reflux disease.

OBJECTIVES: To investigate usefulness, feasibility, and patient satisfaction of an electronic pre-consultation medical history tool (EPMH) in laryngopharyngeal reflux (LPR) work-up. METHODS: Seventy-five patients with LPR were invited to complete electronic medical history assessment prior to laryngology consultation. EPMH collected the following parameters: demographic and epidemiological data, medication, medical and surgical histories, diet habits, stress and symptom findings. Stress and symptoms were assessed with perceived stress scale and reflux symptom score. Duration of consultation, acceptance, and satisfaction of patients (feasibility, usefulness, effectiveness, understanding of questions) were evaluated through a 9-item patient-reported outcome questionnaire. RESULTS: Seventy patients completed the evaluation (93% participation rate). The mean age of cohort was 51.2 ± 15.6 years old. There were 35 females and 35 males. Patients who refused to participate (N = 5) were > 65 years old. The consultation duration was significantly lower in patients who used the EPMH (11.3 ± 2.7 min) compared with a control group (18.1 ± 5.1 min; p = 0.001). Ninety percent of patients were satisfied about EPMH easiness and usefulness, while 97.1% thought that EPMH may improve the disease management. Patients would recommend similar approach for otolaryngological or other specialty consultations in 98.6% and 92.8% of cases, respectively. CONCLUSION: The use of EPMH is associated with adequate usefulness, feasibility, and satisfaction outcomes in patients with LPR. This software is a preliminary step in the development of an AI-based diagnostic decision support tool to help laryngologists in their daily practice. Future randomized controlled studies are needed to investigate the gain of similar approaches on the traditional consultation format.

Moving Forward with Dysphagia Care: Implementing Strategies during the COVID-19 Pandemic and Beyond.

Growing numbers of SARS-CoV-2 cases coupled with limited understanding of transmissibility and virulence, have challenged the current workflow and clinical care pathways for the dysphagia provider. At the same time, the need for non-COVID-19-related dysphagia care persists. Increased awareness of asymptomatic virus carriers and variable expression of the disease have also focused attention to appropriate patient care in the context of protection for the healthcare workforce. The objective of this review was to create a clinical algorithm and reference for dysphagia clinicians across clinical settings to minimize spread of COVID-19 cases while providing optimal care to patients suffering from swallowing disorders. Every practitioner and healthcare system will likely have different constraints or preferences leading to the utilization of one technique over another. Knowledge about this pandemic increases every day, but the algorithms provided here will help in considering the best options for proceeding with safe and effective dysphagia care in this new era.

A Rare Case of Pneumomediastinum after Eustachian Tube Dilation.

Eustachian tube dilation (ETD) is a relatively new intervention for the treatment of eustachian tube dysfunction. Though it previously had no assigned billing code, the American Medical Association recently accepted a new Category I Current Procedural Terminology code application for ETD to be effective in January 2021. Reported complications are uncommon and usually minor. Herein, we present a rare case of massive pneumomediastinum following this procedure. Such major complications are critical to report as ETD becomes a more commonly practiced procedure.

The Relationship Between Hiatal Hernia and Cricopharyngeus Muscle Dysfunction.

Although the precise etiology of cricopharyngeus muscle (CPM) dysfunction (CPMD) is uncertain, many have hypothesized that a hypertrophied CPM may develop as a protective compensation against gastroesophageal reflux disease (GERD). The purpose of this investigation was to evaluate the association between CPMD and the presence of hiatal hernia (HH) in an attempt to elucidate the potential etiology of CPMD. The charts of individuals who underwent video fluoroscopic esophagrams between 01/01/14 and 10/30/16 were reviewed from an electronic database. A group of 50 subjects with reported HH were identified and age- and gender-matched to an individual without HH. The prevalence of CPMD was compared between groups. The mean (± SD) age of the HH cohort was 64 (± 13.4) years and 64 (± 12.8) years for the group without HH (p > 0.05). Thirty-eight percent was male. The frequency of CPMD in the HH group was 78% versus 58% in the non-HH group (p < 0.05). Individuals with a HH were 2.57 times more likely to have evidence of CPMD (95% CI 1.07- 6.15). Although these data suggest an association between GERD and CPMD, further research is required before a causal relationship can be presumed.

Increased mitochondrial energy efficiency in skeletal muscle after long-term fasting: its relevance to animal performance.

In the final stage of fasting, skeletal muscle mass and protein content drastically decrease when the maintenance of efficient locomotor activity becomes crucial for animals to reactivate feeding behaviour and survive a very long period of starvation. As mitochondrial metabolism represents the main physiological link between the endogenous energy store and animal performance, the aim of this study was to determine how a very long, natural period of fasting affected skeletal muscle mitochondrial bioenergetics in king penguin (Aptenodytes patagonicus) chicks. Rates of mitochondrial oxidative phosphorylation were measured in pectoralis permeabilized fibres and isolated mitochondria. Mitochondrial ATP synthesis efficiency and the activities of respiratory chain complexes were measured in mitochondria isolated from pectoralis muscle. Results from long-term (4-5 months) naturally fasted chicks were compared with those from short-term (10 day) fasted birds. The respiratory activities of muscle fibres and isolated mitochondria were reduced by 60% and 45%, respectively, on average in long-term fasted chicks compared with short-term fasted birds. Oxidative capacity and mitochondrial content of pectoralis muscle were lowered by long-term fasting. Bioenergetic analysis of pectoralis muscle also revealed that mitochondria were, on average, 25% more energy efficient in the final stage of fasting (4-5 months) than after 10 days of fasting (short-term fasted birds). These results suggest that the strong reduction in respiratory capacity of pectoralis muscle was partly alleviated by increased mitochondrial ATP synthesis efficiency. Such oxidative phosphorylation optimization can impact animal performance, e.g. the metabolic cost of locomotion or the foraging efficiency.

Applications of ChatGPT in Otolaryngology-Head Neck Surgery: A State of the Art Review.

OBJECTIVE: To review the current literature on the application, accuracy, and performance of Chatbot Generative Pre-Trained Transformer (ChatGPT) in Otolaryngology-Head and Neck Surgery. DATA SOURCES: PubMED, Cochrane Library, and Scopus. REVIEW METHODS: A comprehensive review of the literature on the applications of ChatGPT in otolaryngology was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-analyses statement. CONCLUSIONS: ChatGPT provides imperfect patient information or general knowledge related to diseases found in Otolaryngology-Head and Neck Surgery. In clinical practice, despite suboptimal performance, studies reported that the model is more accurate in providing diagnoses, than in suggesting the most adequate additional examinations and treatments related to clinical vignettes or real clinical cases. ChatGPT has been used as an adjunct tool to improve scientific reports (referencing, spelling correction), to elaborate study protocols, or to take student or resident exams reporting several levels of accuracy. The stability of ChatGPT responses throughout repeated questions appeared high but many studies reported some hallucination events, particularly in providing scientific references. IMPLICATIONS FOR PRACTICE: To date, most applications of ChatGPT are limited in generating disease or treatment information, and in the improvement of the management of clinical cases. The lack of comparison of ChatGPT performance with other large language models is the main limitation of the current research. Its ability to analyze clinical images has not yet been investigated in otolaryngology although upper airway tract or ear images are an important step in the diagnosis of most common ear, nose, and throat conditions. This review may help otolaryngologists to conceive new applications in further research.

Artificial Intelligence in Otolaryngology: Topics in Epistemology & Ethics.

To fuel artificial intelligence (AI) potential in clinical practice in otolaryngology, researchers must understand its epistemic limitations, which are tightly linked to ethical dilemmas requiring careful consideration. AI tools are fundamentally opaque systems, though there are methods to increase explainability and transparency. Reproducibility and replicability limitations can be overcomed by sharing computing code, raw data, and data processing methodology. The risk of bias can be mitigated via algorithmic auditing, careful consideration of the training data, and advocating for a diverse AI workforce to promote algorithmic pluralism, reflecting our population's diverse values and preferences.

Machine learning in the evaluation of voice and swallowing in the head and neck cancer patient.

PURPOSE OF REVIEW: The purpose of this review is to present recent advances and limitations in machine learning applied to the evaluation of speech, voice, and swallowing in head and neck cancer. RECENT FINDINGS: Novel machine learning models incorporating diverse data modalities with improved discriminatory capabilities have been developed for predicting toxicities following head and neck cancer therapy, including dysphagia, dysphonia, xerostomia, and weight loss as well as guiding treatment planning. Machine learning has been applied to the care of posttreatment voice and swallowing dysfunction by offering objective and standardized assessments and aiding innovative technologies for functional restoration. Voice and speech are also being utilized in machine learning algorithms to screen laryngeal cancer. SUMMARY: Machine learning has the potential to help optimize, assess, predict, and rehabilitate voice and swallowing function in head and neck cancer patients as well as aid in cancer screening. However, existing studies are limited by the lack of sufficient external validation and generalizability, insufficient transparency and reproducibility, and no clear superior predictive modeling strategies. Algorithms and applications will need to be trained on large multiinstitutional data sets, incorporate sociodemographic data to reduce bias, and achieve validation through clinical trials for optimal performance and utility.

Concurrent Validity of a Low-Cost Manometer for Objective Assessments of Respiratory Muscle Strength.

OBJECTIVE(S): This study examined the agreement in maximal expiratory (MEP) and inspiratory (MIP) pressure readings between two digital manometers: (1) the MicroRPM - the gold-standard manometer for respiratory muscle strength testing; and (2) the LDM - a low-cost, commercially available, alternative manometer. METHODS: Positive (MEP) and negative (MIP) pressures were simultaneously applied to the MicroRPM and LDM using a 3-liter syringe within a controlled laboratory setting. Pressure readings were compared, and agreement was analyzed using Lin's concordance correlation (ρc ). Agreement was interpreted as 'poor' if <0.90, 'moderate' if 0.90 - <0.95, 'substantial' if 0.95 - <0.99, and 'excellent' if ≥0.99. Twenty percent of the pressure trials were repeated by a second researcher to examine test-retest reliability. RESULTS: A total of 150 trials were completed, ranging from -167 to +208 cmH2 O. There was a median absolute difference of 0.3 cmH2 O in pressure readings between the MicroRPM and the LDM. Lin's concordance correlation revealed 'excellent' agreement between the LDM and MicroRPM devices, with test-retest reliability assessment revealing 'substantial-to-excellent' agreement between the LDM and MicroRPM devices, with a concordance correlation coefficient of ρc = 0.999 (95% CI: 0.999-0.999). CONCLUSIONS: There was a median difference of 1.0% in MEP and MIP pressure readings consistently observed between the LDM and MicroRPM. Despite these relatively small differences, excellent agreement between the two manometers was present. These data suggest the LDM may be a valid, lower cost alternative to the MicroRPM for objectively assessing respiratory strength in clinical practice; however, additional research is needed in healthy adults and in patient populations. LEVEL OF EVIDENCE: NA Laryngoscope, 134:1831-1836, 2024.

The Readability of Patient-Facing Social Media Posts on Common Otolaryngologic Diagnoses.

OBJECTIVE: To assess the readability of patient-facing educational information about the most common otolaryngology diagnoses on popular social media platforms. STUDY DESIGN: Cross-sectional study. SETTING: Social media platforms. METHODS: The top 5 otolaryngologic diagnoses were identified from the National Ambulatory Medical Care Survey Database. Facebook, Twitter, TikTok, and Instagram were searched using these terms, and the top 25 patient-facing posts from unique accounts for each search term and poster type (otolaryngologist, other medical professional, layperson) were identified. Captions, text, and audio from images and video, and linked articles were extracted. The readability of each post element was calculated with multiple readability formulae. Readability was summarized and was compared between poster types, platforms, and search terms via Kruskal-Wallis testing. RESULTS: Median readability, by grade level, by grade level, was greater than 10 for captions, 5 for image-associated text, and 9 for linked articles. Captions and images in posts by laypeople were significantly more readable than captions by otolaryngologists or other medical professionals, but there was no difference for linked articles. All post components were more readable in posts about cerumen than those about other search terms. CONCLUSIONS: When examining the readability of posts on social media regarding the most common otolaryngology diagnoses, we found that many posts are less readable than recommended for patients, and found that posts by laypeople were significantly more readable than those by medical professionals. Medical professionals should work to make educational social media posts more readable to facilitate patient comprehension.

Validation of a Low-Cost Manometer to Assess of Tongue, Lip, Cheek, and Respiratory Strength: A Laboratory-Based Study.

OBJECTIVE(S): The objective of this study was to characterize the level of agreement between three manometers: (1) Iowa Oral Performance Instrument (IOPI)-the reference standard for tongue, lip, and cheek strength assessments; (2) MicroRPM Respiratory Pressure Meter (MicroRPM)-the reference standard for respiratory strength assessments; and (3) Digital Pressure Manometer (DPM)-an alternative, low-cost pressure testing manometer. METHODS: Manual pressures were simultaneously applied to the IOPI and DPM, and to the MicroRPM and DPM, within a controlled laboratory setting. Agreement in pressure readings were analyzed using descriptive statistics, Lin's concordance correlation, and Bland-Altman Plots. Agreement was interpreted as "poor" if ρc < 0.90, "moderate" if ρc = 0.90 - < 0.95, "substantial" if ρc = 0.95 - < 0.99, and "excellent" if ρc ≥ 0.99. RESULTS: Differences in pressure readings between the DPM and clinical reference standards were consistently present yet highly predictable. There was a median absolute difference of 2.0-3.9 kPa between the IOPI and DPM, and 4.5-9.8 cm H2O between the MicroRPM and DPM. Lin's concordance revealed "substantial" agreement between the IOPI and DPM (ρc = 0.98) and the MicroRPM and DPM (ρc = 0.99). CONCLUSION: The DPM revealed higher pressure readings when compared to the IOPI and MicroRPM. However, differences in pressure readings were relatively small, highly predictable, and yielded substantial overall agreement. These findings suggest the DPM may be a valid, lower-cost alternative for objective assessments of tongue, lip, cheek, and respiratory muscle strength. Future research should expand on the present findings in clinical patient populations. LEVEL OF EVIDENCE: NA Laryngoscope, 2024.

New developments in the application of artificial intelligence to laryngology.

PURPOSE OF REVIEW: The purpose of this review is to summarize the existing literature on artificial intelligence technology utilization in laryngology, highlighting recent advances and current barriers to implementation. RECENT FINDINGS: The volume of publications studying applications of artificial intelligence in laryngology has rapidly increased, demonstrating a strong interest in utilizing this technology. Vocal biomarkers for disease screening, deep learning analysis of videolaryngoscopy for lesion identification, and auto-segmentation of videofluoroscopy for detection of aspiration are a few of the new ways in which artificial intelligence is poised to transform clinical care in laryngology. Increasing collaboration is ongoing to establish guidelines and standards for the field to ensure generalizability. SUMMARY: Artificial intelligence tools have the potential to greatly advance laryngology care by creating novel screening methods, improving how data-heavy diagnostics of laryngology are analyzed, and standardizing outcome measures. However, physician and patient trust in artificial intelligence must improve for the technology to be successfully implemented. Additionally, most existing studies lack large and diverse datasets, external validation, and consistent ground-truth references necessary to produce generalizable results. Collaborative, large-scale studies will fuel technological innovation and bring artificial intelligence to the forefront of patient care in laryngology.

Access to Otolaryngologic Telemedicine Care Across the COVID-19 Pandemic at an Urban Tertiary Hospital System.

Objective: To describe associations between patients' demographic characteristics and access to telemedicine services in an urban tertiary academic medical system across the COVID-19 pandemic, and to identify potential barriers to access. Methods: This was a retrospective cohort study conducted at a single-center tertiary academic medical center. The study included adult patients undergoing outpatient otolaryngologic care in person or via telemedicine during 8 week timeframes: before the pandemic, at the onset of the pandemic, and during later parts of the pandemic. Patients were characterized by age, sex, race, insurance type, primary language, portal activation status, income estimate, and visit type. Where appropriate, chi-squared tests, Wilcoxon signed-rank tests, and logistic regression were used to compare demographic factors between the cohorts. Results: A total of 14,240 unique patients [median age, 58 years (range, 18-107 years); 56.5% were female] resulting in a total of 29,457 visits (94.8% in-person and 5.2% telemedicine) were analyzed. Patients seen in person were older than those using telemedicine. Telemedicine visits included a higher proportion of patients with private insurance, and fewer patients with government or no insurance compared to in-person visits. Race, income, and English as primary language were not found to have a significant effect on telemedicine use. Conclusion: In an urban tertiary medical center, we found significant differences in sociodemographic characteristics between patients who accessed otolaryngologic care in person versus via telemedicine through different phases of the COVID pandemic, reflecting possible barriers to care associated with telemedicine. Further studies are needed to develop interventions to improve access.

Female Surgical Ergonomics in Otolaryngology: A National Survey Study.

OBJECTIVE: To assess the nature and impact surgical ergonomic challenges experienced by female otolaryngologists. STUDY DESIGN: National survey study. SETTING: Female otolaryngology residents, fellows and attendings recruited via social media posting and email distribution. METHODS: We distributed a survey study to female otolaryngologists throughout the United States. The height and glove size of participants reporting difficulties with equipment and instruments were compared to those not reporting difficulties. RESULTS: Ninety-six female otolaryngologists participated in our study, comprised of 43% residents, 10% fellows, and 47% attendings. Ninety percent of participants reported difficulties using equipment and 77% of participants reported difficulty with instruments, the most common being nasal endoscopic instruments (28%). The vast majority of participants reported pain during and (or) after the operation (92%). Head and neck (53%) and rhinology (44%) were identified as particularly challenging specialities, but only 25% of participants reported that ergonomics affected their career plans. Participants felt that adjustable equipment (60%), a variety of sizes of instruments (43%), and more discussion around ergonomics (47%) would help. Respondents reported adjusting the operating room to accommodate their size took extra time (44%) and was a mental burden (39%). Participants reporting difficulties with operating room equipment were significantly shorter than those without difficulties (64 inches vs 67 inches, P = .037), and those reporting difficulties with instruments had a smaller median glove size (6 vs 6.5, P = .018). CONCLUSION: Surgical ergonomics represent a challenge for female otolaryngologists, particularly those with smaller hands and shorter height. Partnering with industry, we must address the needs of an increasingly diverse workforce to ensure that all surgeons can operate effectively and comfortably.

Correlations Between Anxiety and/or Depression Diagnoses and Dysphagia Severity.

OBJECTIVE: An increased prevalence of mood and anxiety disorders in patients with dysphagia has been noted previously, but whether dysphagia severity may be exacerbated by anxiety and depression has never been studied before. The purpose of this study is to identify the effect of pre-existing diagnosis of anxiety and/or depression (anxiety/depression) on the Eating Assessment Tool (EAT-10), a validated patient-reported outcome measure for dysphagia. We hypothesized that patients with dysphagia and normal instrumental evaluation have higher EAT-10 score in the presence of pre-existing anxiety and depression. METHODS: A retrospective chart review was conducted of patients seen at the multi-disciplinary dysphagia clinic of an urban academic institution. EAT-10 scores and pre-existing diagnoses of anxiety/depression were collected at the first visit with laryngologists. The two-sample t-test was used to compare mean EAT-10 scores between the anxiety/depression and no anxiety/depression groups, stratified by swallowing dysfunction etiology. RESULTS: The study included 290 consecutive patients seen starting in January 2018. In this cohort, 60 (21%) had pre-existing anxiety, 49 (17%) depression, and 36 (12%) both. Overall, 59 patients had normal swallowing based on instrumental swallowing testing (flexible endoscopic evaluation of swallowing, videofluoroscopic swallow study, esophagram, or esophagoscopy). Among those, mean EAT-10 score was significantly higher in patients with anxiety and/or depression (n = 30) (14.63, SD = 11.42) compared to those with no anxiety and/or depression (n = 29) (8.93, SD = 6.59) (p = 0.023). CONCLUSION: While anxiety/depression may aggravate dysphagia in patients with normal swallowing function, this correlation may not hold in those with objective swallowing dysfunction. LEVEL OF EVIDENCE: 4 Laryngoscope, 134:2115-2120, 2024.

Quantification of Automatic Speech Recognition System Performance on d/Deaf and Hard of Hearing Speech.

OBJECTIVE: To evaluate the performance of commercial automatic speech recognition (ASR) systems on d/Deaf and hard-of-hearing (d/Dhh) speech. METHODS: A corpus containing 850 audio files of d/Dhh and normal hearing (NH) speech from the University of Memphis Speech Perception Assessment Laboratory was tested on four speech-to-text application program interfaces (APIs): Amazon Web Services, Microsoft Azure, Google Chirp, and OpenAI Whisper. We quantified the Word Error Rate (WER) of API transcriptions for 24 d/Dhh and nine NH participants and performed subgroup analysis by speech intelligibility classification (SIC), hearing loss (HL) onset, and primary communication mode. RESULTS: Mean WER averaged across APIs was 10 times higher for the d/Dhh group (52.6%) than the NH group (5.0%). APIs performed significantly worse for "low" and "medium" SIC (85.9% and 46.6% WER, respectively) as compared to "high" SIC group (9.5% WER, comparable to NH group). APIs performed significantly worse for speakers with prelingual HL relative to postlingual HL (80.5% and 37.1% WER, respectively). APIs performed significantly worse for speakers primarily communicating with sign language (70.2% WER) relative to speakers with both oral and sign language communication (51.5%) or oral communication only (19.7%). CONCLUSION: Commercial ASR systems underperform for d/Dhh individuals, especially those with "low" and "medium" SIC, prelingual onset of HL, and sign language as primary communication mode. This contrasts with Big Tech companies' promises of accessibility, indicating the need for ASR systems ethically trained on heterogeneous d/Dhh speech data. LEVEL OF EVIDENCE: 3 Laryngoscope, 2024.

Trust in Machine Learning Driven Clinical Decision Support Tools Among Otolaryngologists.

BACKGROUND: Machine learning driven clinical decision support tools (ML-CDST) are on the verge of being integrated into clinical settings, including in Otolaryngology-Head & Neck Surgery. In this study, we investigated whether such CDST may influence otolaryngologists' diagnostic judgement. METHODS: Otolaryngologists were recruited virtually across the United States for this experiment on human-AI interaction. Participants were shown 12 different video-stroboscopic exams from patients with previously diagnosed laryngopharyngeal reflux or vocal fold paresis and asked to determine the presence of disease. They were then exposed to a random diagnosis purportedly resulting from an ML-CDST and given the opportunity to revise their diagnosis. The ML-CDST output was presented with no explanation, a general explanation, or a specific explanation of its logic. The ML-CDST impact on diagnostic judgement was assessed with McNemar's test. RESULTS: Forty-five participants were recruited. When participants reported less confidence (268 observations), they were significantly (p = 0.001) more likely to change their diagnostic judgement after exposure to ML-CDST output compared to when they reported more confidence (238 observations). Participants were more likely to change their diagnostic judgement when presented with a specific explanation of the CDST logic (p = 0.048). CONCLUSIONS: Our study suggests that otolaryngologists are susceptible to accepting ML-CDST diagnostic recommendations, especially when less confident. Otolaryngologists' trust in ML-CDST output is increased when accompanied with a specific explanation of its logic. LEVEL OF EVIDENCE: 2 Laryngoscope, 134:2799-2804, 2024.