遅発性内リンパ水腫

OBJECTIVE: To evaluate GPT-5 for grading cochlear and vestibular endolymphatic hydrops (EH) on delayed gadolinium-enhanced 3D-FLAIR MRI in Ménière's disease (MD) and to assess prompting strategies and human-AI collaboration. MATERIALS AND METHODS: This retrospective study included 436 patients with MD (872 ears). GPT-5 graded cochlear and vestibular EH under multiple prompting strategies and was compared with junior and senior neuroradiologists in independent and GPT-5-assisted workflows. Accuracy (ACC), AUC, F1 score, agreement, and expert-rated clinical usefulness were assessed. RESULTS: Under basic prompting strategies without structured image descriptions, diagnostic performance remained poor: cochlear hydrops grading ACC reached only 25%, and vestibular hydrops grading ACC reached only 39%, despite the inclusion of clinical history or few-shot examples. In contrast, with enhanced prompting that integrated structured image descriptions, clinical context, and representative examples, GPT-5 showed substantially improved performance, achieving accuracies of 87% for cochlear grading and 80% for vestibular grading, with improved agreement and favorable expert ratings. In human-AI interaction experiments, the Human-first mode improved junior physicians' performance, increasing cochlear grading ACC from 36% to 59% and vestibular grading ACC from 42% to 53%. CONCLUSION: GPT-5 demonstrated limited capability to independently grade EH on MRI when relying primarily on image input. Acceptable diagnostic performance was achieved only after incorporating structured image descriptions, clinical context and representative examples. Under these enhanced prompting conditions, GPT-5 showed potential as an assistive tool for human-AI collaborative interpretation, particularly for less experienced physicians, rather than as a fully autonomous diagnostic system. KEY POINTS: Question Accurate grading of endolymphatic hydrops on 3D-FLAIR MRI is critical for managing Ménière's disease, but current interpretation lacks standardization and efficiency. Findings GPT-5 showed poor EH MRI grading accuracy without image descriptions, while optimized prompting and human-AI collaboration markedly improved performance toward expert interpretation. Clinical relevance Acceptable MRI grading of endolymphatic hydrops by GPT-5 requires structured image descriptions and contextual prompting. This supports its role as a human-AI collaborative assistive tool, not an autonomous system, potentially reducing reliance on limited medical resources.

OBJECTIVE: Delayed endolymphatic hydrops (DEH) and Ménière's disease (MD) are both hydropic ear disease. In MD patients, the endotypes of MD are classified based on pathological types of the endolymphatic sac (ES), which can be distinguished by a radiological indicator: the angular trajectory of the vestibular aqueduct (ATVA). Studies have exhibited that MD patients of different pathological types of ES manifest different clinical characteristics. Here, we aimed to investigate the prevalence of each ATVA category in DEH and MD, and to further explore the clinical significance of ATVA in ipsilateral DEH patients. METHODS: Twenty-seven ipsilateral DEH patients, 114 unilateral MD patients and 45 control participants were enrolled in the present study. The ATVA was determined using temporal bone CT imaging. Clinical features, including gender, age and the caloric test results, were collected. The prevalence of ATVA categories were compared between patients with ipsilateral DEH and unilateral MD. The association between ATVA and clinical features was also examined. RESULTS: (1) Compared to MD-affected ears, ATVA ≤ 120° was more frequent (χ2 = 9.231, p = 0.002), whereas ATVA ≥ 140° was less frequent in DEH-affected ears (χ2 = 6.865, p = 0.009). (2) Compared to control ears, the distribution of ATVA categories in affected ears of ipsilateral DEH patients was not significantly different (χ2 = 1.190, p = 0.510). (3) The prevalence of ATVA ≥ 140°, 120° < ATVA < 140° and ATVA ≤ 120° in affected ears of ipsilateral DEH patients was comparable to that in non-affected ears (χ2 = 1.083, p = 0.568). CONCLUSION: Compared to unilateral MD patients, degenerative ES may be the dominant pathological type in ipsilateral DEH patients due to the high occurrence of ATVA ≤ 120°.

OBJECTIVE: Endolymphatic hydrops (EH), Menière's disease (MD) pathological hallmark, is visualized via 3 T delayed phase MRI. This study examines MRI correlations between EH severity and migraine brain biomarkers. METHODS: Retrospective review of 144 inner ear MRI scans yielded 108 patients with confirmed EH. Two blinded radiologists graded EH, assessed migraine markers (white matter lesions (WML), frontal predominance of WML, enlarged perivascular spaces), and measured cranial nerve volumes. Binary logistic regression tested associations. RESULTS: Mean age 53.6 ± 15.8 years (56.5% female). Grade 2 vestibular EH predominated bilaterally; grade 2 cochlear EH affected 34-43%. No EH associations with migraine biomarkers or cranial nerve volumes were found (all p > 0.05). Left vestibular EH grade 1 linked to sudden hearing loss (OR 5.205, p = 0.018); left cochlear EH grade 1 to aural fullness (OR 5.019, p = 0.016). CONCLUSION: EH severity was not associated with established migraine MRI biomarkers in this cohort. These findings support the concept that EH and migraine-related structural brain changes may represent distinct imaging phenomena.

HYPOTHESIS: Intratympanic (IT) application of betahistine diffuses to the inner ear after 15 minutes of exposure and exhibits no ototoxicity. BACKGROUND: Management of Menière disease involves oral treatment with betahistine; however, is mitigated by a delayed bioavailability (enteral resorption; first-pass effect). Topical IT administration might achieve higher intracochlear concentration. METHODS: Animal study on wild-type rats with different IT-betahistine concentrations: 100 mg/mL, 50 mg/mL, 10 mg/mL, 1 mg/mL, and control (saline). For diffusion (n=18) perilymph was sampled postinjection and tested for betahistine concentration via ultra-high performance liquid chromatography (UHPLC). To assess safety (n=16) each animal was IT-injected every other week for a total of 3 injections to one ear with betahistine and contralateral with saline. Pre- and postinterventional audiometric (otoacoustic emissions; auditory brainstem response, 4 to 32 kHz), and vestibular behavioral testing (forced swim test; beam test) were performed. Cochleae were harvested 2 weeks postinjection for morphologic and histologic processing. RESULTS: UHPLC confirmed dose-dependent diffusion of IT-administered betahistine to the inner ear ( P <0.001). Pre- and postinterventional audiometric thresholds were similar, as well as forced swim and beam test. Morphologic and histologic assessment of cochleae demonstrated integrity of hair cells. CONCLUSION: This preclinical study demonstrated that IT-administered betahistine diffuses dose-dependently into the inner ear and is safe pertaining to cochlear and vestibular function and inner ear microstructures. Higher intracochlear concentration can be achieved by this route of application. This opens the option for a novel treatment for patients with Menière disease.

BACKGROUND: Unilateral Ménière's disease (MD) is characterized by severe episodic attacks of rotational vertigo and audiovestibular signs that include fluctuating sensorineural hearing loss, aural fullness, tinnitus, and vestibular loss. PURPOSE: To investigate the relationship between the degree of audiovestibular loss, symptoms, and the number of vertigo attacks during the previous 6 months and 1 month of patients with MD. STUDY SAMPLE: Sixty patients with active MD refractory to oral treatment. DATA COLLECTION AND ANALYSIS: Hearing (pure-tone audiometry and speech discrimination) and vestibular (caloric test, vestibular evoked myogenic potential [VEMP]) tests were performed. Symptoms were scored using the following validated questionnaires: Vertigo Symptom Scale-short form, Dizziness Handicap Inventory, Tinnitus Handicap Inventory, Aural Fullness Scale, and Functional Level Scale. Linear regression, a factor analysis, and a correlation analysis were conducted. RESULTS: Patients experienced a mean number of 16 attacks (range, 2-65) during the previous 6 months and five attacks (range, 0-30) during the previous 1 month. Hearing and vestibular functions were reduced in the affected ear. Linear regression, the factor analysis, and the correlation analysis showed no association between the degree of audiovestibular loss and the number of vertigo attacks during the previous 6 months or 1 month or symptoms of disease activity. An exploratory analysis showed evidence of some associations between the degree of hearing loss, caloric paresis, and VEMP asymmetry, indicating a consistent pattern of loss across the cochlear semi-circular canals and saccules. CONCLUSION: Although audiovestibular tests help establish diagnoses, they cannot determine current disease activity of unilateral refractory MD. CLINICAL RELEVANCE STATEMENT: It may be prudent to assume that patients with MD and high remaining vestibular function in the affected ear would experience more frequent episodes of vertigo and intense dizziness and experience more troublesome symptoms. Conversely, it may be prudent to assume that patients with MD with low remaining vestibular function in the affected ear would experience fewer episodes of vertigo; however, as shown by these results, this does not seem to be the case. No obvious audiovestibular markers suggest that intratympanic therapy for vertigo treatment should be delayed.