Developing an Artificial Intelligence Tool to Predict Vocal Cord Pathology in Primary Care Settings.

OBJECTIVES: Diagnostic tools for voice disorders are lacking for primary care physicians. Artificial intelligence (AI) tools may add to the armamentarium for physicians, decreasing the time to diagnosis and limiting the burden of dysphonia. METHODS: Voice recordings of patients were collected from 2019 to 2021 using smartphones. The Saarbruecken dataset was included for comparison. Audio files were converted to mel-spectrograms using TensorFlow. Diagnostic categories were created to group pathology, including neurological and muscular disorders, inflammatory, mass lesions, and normal. The samples were further separated into sustained/a/and the rainbow passage. RESULTS: Two hundred three prospective samples and 1131 samples were used from the Saarbruecken database. The AI detected abnormal pathology with an F1-score of 98%. The artificial neural network (ANN) differentiated key pathologies, including unilateral paralysis, laryngitis, adductor spasmodic dysphonia (ADSD), mass lesions, and normal samples with 39%-87% F-1 scores. The Calgary database models had higher F-1 scores in a head-to-head comparison to the Saarbruecken and combined datasets (87% vs. 58% and 50%). The AI outperformed otolaryngologists using a standardized test set of recordings (83% compared to 55% ± 15%). CONCLUSION: An AI tool was created to differentiate pathology by individual or categorical diagnosis with high evaluation metrics. Prospective data should be collected in a controlled fashion to reduce intrinsic variability between recordings. Multi-center data collaborations are imperative to increase the prediction capability of AI tools for detecting vocal cord pathology. We provide proof-of-concept for an AI tool to assist primary care physicians in managing dysphonic patients. LEVEL OF EVIDENCE: 3 Laryngoscope, 133:1952-1960, 2023.

Otoscopic diagnosis using computer vision: An automated machine learning approach.

OBJECTIVE: Access to otolaryngology is limited by lengthy wait lists and lack of specialists, especially in rural and remote areas. The objective of this study was to use an automated machine learning approach to build a computer vision algorithm for otoscopic diagnosis capable of greater accuracy than trained physicians. This algorithm could be used by primary care providers to facilitate timely referral, triage, and effective treatment. METHODS: Otoscopic images were obtained from Google Images (Google Inc., Mountain View, CA), from open access repositories, and within otolaryngology clinics associated with our institution. After preprocessing, 1,366 unique images were uploaded to the Google Cloud Vision AutoML platform (Google Inc.) and annotated with one or more of 14 otologic diagnoses. A consensus set of labels for each otoscopic image was attained, and a multilabel classifier architecture algorithm was trained. The performance of the algorithm on an 89-image test set was compared to the performance of physicians from pediatrics, emergency medicine, otolaryngology, and family medicine. RESULTS: For all diagnoses combined, the average precision (positive predictive value) of the algorithm was 90.9%, and the average recall (sensitivity) was 86.1%. The algorithm made 79 correct diagnoses with an accuracy of 88.7%. The average physician accuracy was 58.9%. CONCLUSION: We have created a computer vision algorithm using automated machine learning that on average rivals the accuracy of the physicians we tested. Fourteen different otologic diagnoses were analyzed. The field of medicine will be changed dramatically by artificial intelligence within the next few decades, and physicians of all specialties must be prepared to guide that process. LEVEL OF EVIDENCE: NA Laryngoscope, 130:1408-1413, 2020.

Building an Otoscopic screening prototype tool using deep learning.

BACKGROUND: Otologic diseases are often difficult to diagnose accurately for primary care providers. Deep learning methods have been applied with great success in many areas of medicine, often outperforming well trained human observers. The aim of this work was to develop and evaluate an automatic software prototype to identify otologic abnormalities using a deep convolutional neural network. MATERIAL AND METHODS: A database of 734 unique otoscopic images of various ear pathologies, including 63 cerumen impactions, 120 tympanostomy tubes, and 346 normal tympanic membranes were acquired. 80% of the images were used for the training of a convolutional neural network and the remaining 20% were used for algorithm validation. Image augmentation was employed on the training dataset to increase the number of training images. The general network architecture consisted of three convolutional layers plus batch normalization and dropout layers to avoid over fitting. RESULTS: The validation based on 45 datasets not used for model training revealed that the proposed deep convolutional neural network is capable of identifying and differentiating between normal tympanic membranes, tympanostomy tubes, and cerumen impactions with an overall accuracy of 84.4%. CONCLUSION: Our study shows that deep convolutional neural networks hold immense potential as a diagnostic adjunct for otologic disease management.

A Novel Multi-Biomarker Assay for Non-Invasive Quantitative Monitoring of Kidney Injury.

The current standard of care measures for kidney function, proteinuria, and serum creatinine (SCr) are poor predictors of early-stage kidney disease. Measures that can detect chronic kidney disease in its earlier stages are needed to enable therapeutic intervention and reduce adverse outcomes of chronic kidney disease. We have developed the Kidney Injury Test (KIT) and a novel KIT Score based on the composite measurement and validation of multiple biomarkers across a unique set of 397 urine samples. The test is performed on urine samples that require no processing at the site of collection and without target sequencing or amplification. We sought to verify that the pre-defined KIT test, KIT Score, and clinical thresholds correlate with established chronic kidney disease (CKD) and may provide predictive information on early kidney injury status above and beyond proteinuria and renal function measurements alone. Statistical analyses across six DNA, protein, and metabolite markers were performed on a subset of residual spot urine samples with CKD that met assay performance quality controls from patients attending the clinical labs at the University of California, San Francisco (UCSF) as part of an ongoing IRB-approved prospective study. Inclusion criteria included selection of patients with confirmed CKD and normal healthy controls; exclusion criteria included incomplete or missing information for sample classification, logistical delays in transport/processing of urine samples or low sample volume, and acute kidney injury. Multivariate logistic regression of kidney injury status and likelihood ratio statistics were used to assess the contribution of the KIT Score for prediction of kidney injury status and stage of CKD as well as assess the potential contribution of the KIT Score for detection of early-stage CKD above and beyond traditional measures of renal function. Urine samples were processed by a proprietary immunoprobe for measuring cell-free DNA (cfDNA), methylated cfDNA, clusterin, CXCL10, total protein, and creatinine. The KIT Score and stratified KIT Score Risk Group (high versus low) had a sensitivity and specificity for detection of kidney injury status (healthy or CKD) of 97.3% (95% CI: 94.6-99.3%) and 94.1% (95% CI: 82.3-100%). In addition, in patients with normal renal function (estimated glomerular filtration rate (eGFR) ≥ 90), the KIT Score clearly identifies those with predisposing risk factors for CKD, which could not be detected by eGFR or proteinuria (p < 0.001). The KIT Score uncovers a burden of kidney injury that may yet be incompletely recognized, opening the door for earlier detection, intervention and preservation of renal function.

Rhinologic and oral-maxillofacial complications from scuba diving: a systematic review with recommendations.

Rhinologic and oral maxillofacial complications from scuba diving are common, representing approximately 35% of head and neck pathology related to diving. We performed a systematic and comprehensive literature review on the pathophysiology, diagnosis, and treatment of rhinologic and oral maxillofacial pathology related to diving. This included complications due to sinus barotrauma, barodontalgia, odontocrexis, temporomandibular joint dysfunction, partially dentulous patients, and considerations for patients following major head and neck surgery. Of 113 papers accessed, 32 were included in the final synthesis. We created a succinct summary on each topic that should inform clinical decision making by otolaryngologists, dive medicine specialists and primary care providers when faced with pathology of these anatomic sub-sites.

Endoscope-assisted nasal dermoid excision with an open rhinoplasty approach.

Congenital midline nasal masses have been traditionally excised through rhinotomies and bicoronal incisions. These approaches are disfiguring and potentially morbid, leading pediatric otolaryngologists to seek out less invasive, endoscopic-assisted approaches. Here we present a nasal dermoid excised with a minimally invasive, endoscopic assisted open rhinoplasty approach. The preoperative workup as well as perioperative steps for excision are documented, with the hope that this will assist other practitioners with this approach. An endoscopic open rhinoplasty technique affords excellent visualisation, with a superior cosmetic outcome. A preoperative MRI is essential to ensure the lesion is amenable to this approach.

Assessing the performance of a Loop Mediated Isothermal Amplification (LAMP) assay for the detection and subtyping of high-risk suptypes of Human Papilloma Virus (HPV) for Oropharyngeal Squamous Cell Carcinoma (OPSCC) without DNA purification.

BACKGROUND: Oropharyngeal Squamous Cell Carcinoma (OPSCC) is increasing in incidence despite a decline in traditional risk factors. Human Papilloma Virus (HPV), specifically subtypes 16, 18, 31 and 35, has been implicated as the high-risk etiologic agent. HPV positive cancers have a significantly better prognosis than HPV negative cancers of comparable stage, and may benefit from different treatment regimens. Currently, HPV related carcinogenesis is established indirectly through Immunohistochemistry (IHC) staining for p16, a tumour suppressor gene, or polymerase chain reaction (PCR) that directly tests for HPV DNA in biopsied tissue. Loop mediated isothermal amplification (LAMP) is more accurate than IHC, more rapid than PCR and is significantly less costly. In previous work we showed that a subtype specific HPV LAMP assay performed similar to PCR on purified DNA. In this study we examined the performance of this LAMP assay without DNA purification. METHODS: We used LAMP assays using established primers for HPV 16 and 18, and new primers for HPV 31 and 35. LAMP reaction conditions were tested on serial dilutions of plasmid HPV DNA to confirm minimum viral copy number detection thresholds. LAMP was then performed directly on different human cell line samples without DNA purification. RESULTS: Our LAMP assays could detect 105, 103, 104, and 105 copies of plasmid DNA for HPV 16, 18, 31, and 35, respectively. All primer sets were subtype specific, with no cross-amplification. Our LAMP assays also reliably amplified subtype specific HPV DNA from samples without requiring DNA isolation and purification. CONCLUSIONS: The high risk OPSCC HPV subtype specific LAMP primer sets demonstrated, excellent clinically relevant, minimum copy number detection thresholds with an easy readout system. Amplification directly from samples without purification illustrated the robust nature of the assay, and the primers used. This lends further support HPV type specific LAMP assays, and these specific primer sets and assays can be further developed to test for HPV in OPSCC in resource and lab limited settings, or even bedside testing.

Scuba diving and otology: a systematic review with recommendations on diagnosis, treatment and post-operative care.

Scuba diving is a popular recreational and professional activity with inherent risks. Complications related to barotrauma and decompression illness can pose significant morbidity to a diver's hearing and balance systems. The majority of dive-related injuries affect the head and neck, particularly the outer, middle and inner ear. Given the high incidence of otologic complications from diving, an evidence-based approach to the diagnosis and treatment of otic pathology is a necessity. We performed a systematic and comprehensive literature review including the pathophysiology, diagnosis, and treatment of otologic pathology related to diving. This included inner, middle, and outer ear anatomic subsites, as well as facial nerve complications, mal de debarquement syndrome, sea sickness and fitness to dive recommendations following otologic surgery. Sixty-two papers on diving and otologic pathology were included in the final analysis. We created a set of succinct evidence-based recommendations on each topic that should inform clinical decisions by otolaryngologists, dive medicine specialists and primary care providers when faced with diving-related patient pathology.

The role of bone conduction hearing aids in congenital unilateral hearing loss: A systematic review.

OBJECTIVES: To systematically review the literature on the audiological and/or quality of life benefits of a bone conduction hearing aid (BCHA) in children with congenital unilateral conductive or sensorineural deafness. METHODS: A systematic search was performed according to the PRISMA guidelines using the PubMed, Medline, and Embase databases. Data were collected on the following outcomes of interest: speech reception threshold, speech discrimination, sound localization, and quality of life measures. Given the heterogeneity of the data for quantitative analysis, the results are qualitatively summarized. RESULTS: Eight studies were included in the review. Four studies examined the audiological outcomes associated with bone conduction hearing aid implantation. There was a consistent gain in speech reception thresholds and speech discrimination, especially in noisy environments. Results pertaining to sound localization was inconsistent. The studies that examined quality of life measures reported a high usage rate of BCHAs among children. Quality of life improvements are reported with suggested benefit in the subdomain of learning. CONCLUSION: Given the potential benefits of a BCHA, along with the fact that it can be safely trialed using a headband, it is reasonable to trial a BCHA in children with congenital unilateral deafness. Should the trial offer audiological and/or quality of life benefits for the individual child, then BCHA implantation can be considered.

Making a case for high-volume robotic surgery centers: A cost-effectiveness analysis of transoral robotic surgery.

OBJECTIVE: To evaluate the cost-effectiveness of transoral robotic surgery (TORS) compared to intensity-modulated radiotherapy (IMRT) for early stage (T1-2, N0, M0) oropharyngeal squamous cell carcinoma (OPSCC). PATIENTS AND METHODS: A Markov decision tree model with a 5-year time horizon was developed. Comparative groups were: i) TORS with concurrent ipsilateral neck dissection +/- adjunctive IMRT, and ii) primary IMRT. Primary outcome was cost/quality adjusted life year (QALY). Perspective was the United States third party payer. Costs and effects were discounted at a rate of 3.5%. A threshold and probabilistic sensitivity analysis were performed. RESULTS: TORS strategy cost $30,992 and provided 4.81 QALYs/patient. The IMRT strategy cost $26,033 and provided a total of 4.78 QALYs/patient. The incremental cost effectiveness ratio for TORS vs. IMRT in the reference case was $165,300/QALY. The probability that TORS is cost-effective compared to IMRT at a maximum willingness-to-pay threshold of $50,000/QALY is 42%. CONCLUSION: An IMRT strategy for management of early stage OPSCC is more likely to be cost-effective compared to TORS. To improve the value of TORS for early stage OPSCC, consolidating TORS procedures to create high-volume centers of excellence may be a potential strategy to increase incremental effectiveness and reduce incremental costs. J. Surg. Oncol. 2015 111:155-163. © 2015 Wiley Periodicals, Inc.

Silver nitrate mimicking a foreign body in the pharyngeal mucosal space.

Silver nitrate is sometimes used as a means of chemical cauterization for control of minor bleeding and management of hypergranulation tissue following bedside head and neck procedures. There are only few reports available on the imaging appearance of silver nitrate and its potential to mimic a foreign body. We report a case of a patient presenting with dysphagia, odynophagia, and fever following dental work who had a peritonsillar incision and drainage for treatment of a deep neck space infection. During the procedure, silver nitrate was applied to halt the bleeding. Patient was subsequently transferred to another institution. Since the patient was not showing significant clinical improvement on antibiotic therapy, a computed tomography (CT) scan was performed demonstrating a hyperdense structure lodged in the pharyngeal mucosal space in the oropharynx and soft palate that was mistaken for a foreign body such as bone. Silver nitrate can have density similar to bone but does not have the normal architecture of bone with cortex and marrow on CT. Familiarity with the appearance of silver nitrate on CT, lack of bone architecture, and proper documentation and communication of the use of silver nitrate to the consultant radiologist and medical personnel could help avoid misdiagnosis and potentially unnecessary surgical exploration.

Early versus late tracheostomy: a systematic review and meta-analysis.

OBJECTIVE: To investigate whether early tracheostomy leads to improved outcomes compared with late tracheostomy. DATA SOURCES: Ovid MEDLINE (including PubMed), Embase, and the Cochrane Central Register of Controlled Trials. REVIEW METHODS: A systematic search was performed of the above-mentioned databases according to PRISMA guidelines. Data were collected on the following outcomes of interest: hospital mortality, intensive care unit length of stay, length of mechanical ventilation, incidence of pneumonia, laryngotracheal injury, and sedation use. Analysis was performed using the RevMan 5 software (Cochrane Collaboration, Oxford, England). RESULTS: Eleven studies were included for analysis. There was a significant decrease in the intensive care unit length of stay in the early tracheostomy group (weighted mean difference, -9.13 days; 95% confidence interval [CI], -17.55 to -0.70; P = .03). There was no significant difference in hospital mortality (relative risk, 0.84; 95% CI, 0.67 to 1.04; P = .11). A pooled analysis was not performed for the incidence of pneumonia or length of mechanical ventilation, secondary to considerable heterogeneity among the studies. None of the studies reporting laryngotracheal outcomes found a significant difference between the early and late tracheostomy groups, whereas all 3 studies reporting sedation use found a significant decrease in the early tracheostomy group. CONCLUSION: Early tracheostomy performed within 7 days of intubation was associated with a decrease in intensive care unit length of stay. No difference was found in hospital mortality. Insufficient data currently exist to make conclusions about the effect of early tracheostomy on the incidence of pneumonia, length of mechanical ventilation, laryngotracheal injury, or sedation use.

The significance of Streptococcus anginosus group in intracranial complications of pediatric rhinosinusitis.

OBJECTIVE: To assess the significance of the Streptococcus anginosus group in intracranial complications of pediatric patients with rhinosinusitis. DESIGN: Retrospective cohort study. SETTING: Tertiary pediatric hospital. PATIENTS: A 20-year review of medical records identified patients with intracranial complications resulting from rhinosinusitis. In the 50 cases identified, S anginosus was the most commonly implicated bacterial pathogen in 14 (28%). Documented data included demographics, cultured bacteria, immune status, sinuses involved, type of intracranial complication, otolaryngologic surgical and neurosurgical intervention, type and duration of antibiotics used, and resulting neurologic deficits. Complications and outcomes of cases of S anginosus group-associated rhinosinusitis were compared with those of other bacteria. MAIN OUTCOME MEASURES: The severity and outcomes of intracranial complications of pediatric rhinosinusitis due to S anginosus group bacteria compared with other bacteria. RESULTS: Infection caused by the S anginosus group resulted in more severe intracranial complications (P = .001). In addition, patients with S anginosus group-associated infections were more likely to require neurosurgical intervention (P < .001) and develop long-term neurologic deficits (P = .02). Intravenous antibiotics were administered for a longer duration (P < .001) for S anginosus group-associated infections. CONCLUSIONS: Rhinosinusitis associated with the S anginosus group should be considered a more serious infection relative to those caused by other pathogens. Streptococcus anginosus group bacteria are significantly more likely than other bacteria to cause more severe intracranial complications and neurologic deficits and to require neurosurgical intervention. A low threshold for intervention should be used for infection caused by this pathogen.