Citations

ObjectiveThe purpose of this study was to evaluate the performance of advanced large language models from OpenAI (GPT‐3.5 and GPT‐4), Google (PaLM2 and MedPaLM), and an open source model from Meta (Llama3:70b) in answering clinical test multiple choice questions in the field of otolaryngology—head and neck surgery.MethodsA dataset of 4566 otolaryngology questions was used; each model was provided a standardized prompt followed by a question. One hundred questions that were answered incorrectly by all models were further interrogated to gain insight into the causes of incorrect answers.ResultsGPT4 was the most accurate, correctly answering 3520 of 4566 questions (77.1%). MedPaLM correctly answered 3223 of 4566 (70.6%) questions, while llama3:70b, GPT3.5, and PaLM2 were correct on 3052 of 4566 (66.8%), 2672 of 4566 (58.5%), and 2583 of 4566 (56.5%) questions. Three hundred and sixty‐nine questions were answered incorrectly by all models. Prompts to provide reasoning improved accuracy in all models: GPT4 changed from incorrect to correct answer 31% of the time, while GPT3.5, Llama3, PaLM2, and MedPaLM corrected their responses 25%, 18%, 19%, and 17% of the time, respectively.ConclusionLarge language models vary in their understanding of otolaryngology‐specific clinical knowledge. OpenAI's GPT4 has a strong understanding of core concepts as well as detailed information in the field of otolaryngology. Its baseline understanding in this field makes it well‐suited to serve in roles related to head and neck surgery education provided that the appropriate precautions are taken and potential limitations are understood.Level of EvidenceNA Laryngoscope, 135:629–634, 2025

AbstractObjectiveThis study investigated the comparative performance of ear, nose, and throat (ENT) physicians in correctly detecting ear abnormalities when reviewing digital otoscopy imaging using 3 different visualization methods, including computer‐assisted composite images called “SelectStitch,” single video frame “Still” images, and video clips. The study also explored clinicians' diagnostic confidence levels and the time to make a diagnosis.Study DesignClinician diagnostic reader study.SettingOnline diagnostic survey of ENT physicians.MethodsNine ENT physicians reviewed digital otoscopy examinations from 86 ears with various diagnoses (normal, perforation, retraction, middle ear effusion, tympanosclerosis). Otoscopy examinations used artificial‐intelligence (AI)‐based computer‐aided composite image generation from a video clip (SelectStitch), manually selected best still frame from a video clip (Still), or the entire video clip. Statistical analyses included comparisons of ability to detect correct diagnosis, confidence levels, and diagnosis times.ResultsThe ENT physicians' ability to detect ear abnormalities (33.2%‐68.7%) varied depending on the pathologies. SelectStitch and Still images were not statistically different in detecting abnormalities (P > .50), but both were different from Video (P < .01). However, the performance improvement observed with Videos came at the cost of significantly longer time to determining the diagnosis. The level of confidence in the diagnosis was positively associated with correct diagnoses, but varied by particular pathology.ConclusionThis study explores the potential of computer‐assisted techniques like SelectStitch in enhancing otoscopic diagnoses and time‐saving, which could benefit telemedicine settings. Comparable performance between computer‐generated and manually selected images suggests the potential of AI algorithms for otoscopy applications.

AbstractBackground and PurposeTo compare the performance of the photon‐counting detector (PCD)‐CT against a cutting‐edge energy‐integrated detector (EID)‐CT to visualize different segments of the chorda tympani nerve.Materials and MethodsAn IRB‐approved protocol prospectively enrolled patients for temporal bone CT on both EID‐CT and PCD‐CT scanners. Fellowship‐trained neuroradiologists compared visualization of chorda tympani segments on each scanner using a 5‐point Likert scale (1 indicating EID is far superior to PCD, 3 indicating they are equivalent, and 5 indicating PCD is far superior to EID). Scores were averaged across segments and tested using a one‐sample Wilcoxon signed‐rank test with a one‐sided alternative hypothesis.ResultsForty temporal bones were evaluated. Average Likert scores for visualization of chorda tympani segments‐chorda tympani canal/proximal segment, middle ear/posterior tympanic segment, and petrotympanic fissure/distal segment‐ on the PCD compared to EID scanner were 4.2 (SD = 0.4), 4.4 (0.5), and 4 (0.0), respectively. The anterior tympanic segment of the chorda tympani was not well visualized on either platform or therefore not scored. PCD compared to the EID scanner had better overall performance (Median = 4.2, 95% CI: [4.1, 5.0], p‐value < .001).ConclusionPCD‐CT improves visualization of the petrous and intratympanic chorda tympani nerve segments, chorda tympani canaliculus, and petrotympanic fissure compared to comparable EID‐CT examinations. The improved visualization of the chorda tympani is important for the characterization of middle ear pathology involving this structure and presurgical planning to determine the size of the facial recess for cochlear implantation and to estimate the risk of nerve injury.

Objectives Describe a single institution's cochlear implant outcomes for patients with inner ear schwannomas (IES) in the setting of various tumor management strategies (observation, surgical resection, or stereotactic radiosurgery [SRS]). Study Design Single-institution retrospective review. Patients Patients diagnosed with isolated, sporadic IES who underwent cochlear implantation (CI). Interventions CI with or without IES treatment. Main Outcome Measures Speech perception outcomes, tumor status. Results Twelve patients with IES underwent CI with a median audiologic and radiologic follow-up of 12 months. Six patients underwent complete resection of the tumor at the time of CI, four underwent tumor observation, and two underwent SRS before CI. At 1 year after CI for all patients, the median consonant–nucleus–consonant (CNC) word score was 55% (interquartile range, 44–73%), and the median AzBio sentence in quiet score was 77% (interquartile range, 68–93%). Overall, those with surgical resection performed similarly to those with tumor observation (CNC 58 versus 61%; AzBio in quiet 74 versus 91%, respectively). Patients who underwent tumor resection before implantation had a wider range of speech performance outcomes compared with patients who underwent tumor observation. Two patients had SRS treatment before CI (10 months previous and same-day as CI) with CNC word scores of 6 and 40%, respectively. Conclusions Patients with IES who underwent CI demonstrated similar speech performance outcomes (CNC 56% and AzBio 82%), when compared with the general cochlear implant population. Patients who underwent either tumor observation or surgical resection performed well after CI.

AbstractIntroductionSurgical techniques for sellar reconstruction include no reconstruction, use of synthetic materials, autologous grafts, and/or vascularized flaps. The aim of this study was to conduct a multi‐center study comparing the efficacy and postoperative morbidity associated with different sellar reconstruction techniques.MethodsA retrospective chart review of patients who underwent endoscopic transsphenoidal surgery for pituitary tumors from five participating sites between January 2021 and March 2023 was performed. The variables included demographics, tumor characteristics, reconstruction technique, postoperative cerebrospinal fluid leak (CSF) leak, and 22‐item Sino‐Nasal Outcome Test (SNOT‐22) scores. Comparisons of postoperative complications, SNOT‐22 scores, and duration of surgery by type of onlay reconstruction were evaluated using Fisher's exact test, analysis of variance, and Kruskal‒Wallis test.ResultsFive hundred and one patients were identified. The median tumor size was 2.1 cm, and 64% were non‐functioning. Intraoperative CSF leak was identified in 38% of patients. A total of 89% of patients underwent onlay reconstruction: 49% were reconstructed with mucosal grafts, 35% with nasoseptal flaps, and 5% with other onlay techniques. Nasoseptal flaps were utilized more frequently in the setting of giant pituitary adenomas (>3 cm), medial cavernous sinus wall resection, and high‐flow intraoperative CSF leaks. Cases who utilized mucosal grafts had an overall shorter operating time (median: 183 min vs. 240 min; p < 0.001). Five postoperative CSF leaks were identified, and therefore, statistical analysis could not be performed for this complication.ConclusionThe effectiveness and morbidity of different sellar reconstruction techniques are comparable. Vascularized flaps were utilized more frequently in the setting of larger tumors and high‐flow intraoperative CSF leaks.

AbstractObjectiveTo evaluate tumor control and facial nerve outcomes after gross‐total (GTR), near‐total (NTR), and subtotal resection (STR) of sporadic vestibular schwannomas (VS).Data SourcesPubMed, Cochrane Central Register of Controlled Trials, Web of Science, and Scopus databases were searched in August 2021 through inception following PRISMA guidelines.Review MethodsEnglish language articles reporting tumor control and facial nerve outcomes of adults (≥18 years) with NTR and STR of VS were evaluated. Study characteristics, demographics data, tumor characteristics, type of surgical intervention, and outcome measures on tumor control and facial nerve function were collected. Pooled relative risk (RR) estimates for tumor recurrence and facial nerve outcomes were calculated and stratified by extent of resection.ResultsFrom an initial search of 2504 articles, 48 studies were included in the analysis. When comparing 1108 patients who underwent NTR to 3349 patients with GTR, the pooled RR of recurrence in the NTR cohort was 2.94 (95% confidence interval [CI] 1.65‐5.24, P = .0002). When comparing 1016 patients who underwent STR to 6171 patients with GTR, the pooled RR of recurrence in the STR cohort was 11.50 (95% CI 6.64‐19.92, P < .0001). Estimates for risk of tumor regrowth for less‐than‐complete resection are presented. There was no elevated risk of adverse facial nerve outcome (defined as House‐Brackmann grade III and above) in each category of extent of resection compared to GTR.ConclusionExtent of resection predicts risk of tumor recurrence/regrowth following microsurgical resection. Favorable facial nerve outcome should be weighed against the increased risk of regrowth and the potential need for further treatment.

Objective To describe the rare process of osteolytic labyrinthitis, previously referred to as labyrinthine sequestrum, which involves progressive obliteration of the bony and membranous labyrinth with eventual supplantation with soft tissue and, in some cases, bony sequestrum. Patients Three patients with diverse presentations of osteolytic labyrinthitis from two tertiary care academic medical centers. Intervention(s) Case series report analyzing the relevant clinical, radiologic, pathologic, and surgical data on our patients with osteolytic labyrinthitis and comparing these index cases to the existing literature. Main Outcome Measure(s) We describe the varying image findings seen in osteolytic labyrinthitis on computed tomography and magnetic resonance imaging. Also, we report successful surgical intervention and hearing rehabilitation with cochlear implantation in patients with osteolytic labyrinthitis. Results Our three patients presented with profound sudden sensorineural hearing loss and vertigo consistent with labyrinthitis. None of the three patients had a history of chronic otitis media. Imaging workup revealed varying degrees of erosion to the otic capsule bone demonstrating the spectrum of disease seen in osteolytic labyrinthitis. Although two cases showed osteolytic changes to the semicircular canals and vestibule, the first case revealed frank bony sequestrum within the obliterated labyrinth. The three cases were taken for surgical debridement and cochlear implantation. Conclusions We propose the new term, osteolytic labyrinthitis—previously referred to as labyrinthine sequestrum—to describe the rare spectrum of disease characterized by destruction of the osseous and membranous labyrinth and potential supplantation with bony sequestrum. Cochlear implantation is a viable option in selected patients with osteolytic labyrinthitis.

Objective To describe outcomes of patients with sporadic vestibular schwannoma (VS) who underwent repeat stereotactic radiosurgery (SRS) after primary SRS failure. Study Design Multi-institutional historical cohort study. Setting Five tertiary care referral centers. Patients Adults ≥18 years old with sporadic VS. Intervention Primary and repeat treatment with SRS. Main Outcome Measure Microsurgery-free survival after repeat SRS. Results Across institutions, 32 patients underwent repeat SRS after primary SRS. Most patients (74%) had tumors with cerebellopontine angle extension at primary SRS (median size, 13.5 mm [interquartile range, 7.5–18.8] mm). After primary SRS, patients underwent repeat SRS at a median of 4.8 years (interquartile range, 3.2–5.7 yr). For treatment modality, 30 (94%) patients received gamma knife for primary treatment and 31 (97%) patients received gamma knife as their repeat treatment. Median tumor volume increased from 0.970 cm3 at primary SRS to 2.200 cm3 at repeat SRS. Facial nerve function worsened in two patients after primary SRS and in two patients after repeat SRS. There were no instances of intracranial complications after repeat SRS. Microsurgery-free survival rates (95% confidence interval; number still at risk) at 1, 3, and 5 years after repeat SRS were 97% (90–100%, 24), 84% (71–100%, 13), and 68% (48–96%, 6), respectively. There was one occurrence of malignancy diagnosed after repeat radiosurgery. Conclusion Overall, repeat SRS for sporadic VS has comparable risk profile, but lower rates of tumor control, compared with primary SRS.

Objective To examine patient preference after stapedotomy versus cochlear implantation in a unique case of a patient with symmetrical profound mixed hearing loss and similar postoperative speech perception improvement. Patients An adult patient with bilateral symmetrical far advanced otosclerosis, with profound mixed hearing loss. Intervention Stapedotomy in the left ear, cochlear implantation in the right ear. Main Outcome Measure Performance on behavioral audiometry, and subjective report of hearing and intervention preference. Results A patient successfully underwent left stapedotomy and subsequent cochlear implantation on the right side, per patient preference. Preoperative audiometric characteristics were similar between ears (pure-tone average [PTA] [R: 114; L: 113 dB]; word recognition score [WRS]: 22%). Postprocedural audiometry demonstrated significant improvement after stapedotomy (PTA: 59 dB, WRS: 75%) and from cochlear implant (PTA: 20 dB, WRS: 60%). The patient subjectively reported a preference for the cochlear implant ear despite having substantial gains from stapedotomy. A nuanced discussion highlighting potentially overlooked benefits of cochlear implants in far advanced otosclerosis is conducted. Conclusion In comparison with stapedotomy and hearing aids, cochlear implantation generally permits greater access to sound among patients with far advanced otosclerosis. Though the cochlear implant literature mainly focuses on speech perception outcomes, an underappreciated benefit of cochlear implantation is the high likelihood of achieving “normal” sound levels across the audiogram.