Large language models (LLMs) are increasingly applied in clinical decision support, yet their behavior when presented with structured textual descriptions of imaging findings, particularly in salivary gland disease, remains poorly defined. The present study was set out to evaluate how the responses of different LLMs change as the textual input is progressively enriched. The models did not receive imaging data; all inputs were text. This exploratory, simulation-based observational study assessed 3 LLMs (ChatGPT, Gemini, Claude) across 10 salivary gland cases. For each case, a standardized textual description of the imaging findings was prepared and submitted as a text prompt; no image files were uploaded to any platform. Each case was evaluated under 3 input conditions of increasing informational content: single-plane textual description, multiplanar textual description, and multiplanar textual description plus clinical information, generating 90 total observations derived from repeated measurement of the same 10 cases. Primary outcomes were the model-generated severity score, self-reported diagnostic confidence, and the number of additional reported anatomical structures. AI outputs were not benchmarked against a verified reference diagnosis. The primary analysis was a proportional-odds model with input condition as the fixed effect of interest, AI system as a covariate, and cluster-robust standard errors by case; non-parametric tests were used as sensitivity analyses. Model-generated severity scores did not significantly differ among LLMs (χ²=2.31; P = 0.31). Self-reported diagnostic confidence increased significantly as the textual input was enriched (Friedman χ²=18.86; P < 0.001; proportional-odds model, multiplanar description plus clinical information vs. single-plane description, cumulative OR 5.43, 95
Cocaine-induced midline destructive lesions (CIMDL) may cause progressive septonasal and palatal destruction, but the optimal timing of reconstruction remains undefined. Because relapse is common, particularly within the first year, premature surgery may lead to failure and, in advanced cases, severe complications. This work synthesizes the available evidence and expert opinion to guide surgical timing and practical management. We performed a narrative evidence-based review of studies addressing cocaine relapse, abstinence verification, and reconstructive or prosthetic management in CIMDL, integrated with multidisciplinary expert consensus. Key relapse data were drawn from large longitudinal cohorts, including national treatment and follow-up studies and analyses of continuing care and initial abstinence. Relapse risk remains clinically significant throughout the first 12 months after treatment, with weekly-use relapse rates of approximately 20–24
Riedel thyroiditis (RT) is a rare fibroinflammatory disorder that can mimic malignancy clinically and radiologically. We report an unusually extensive, ultimately fatal case and highlight its diagnostic and therapeutic lessons. Case report of a 40-year-old man, with review of the clinical, imaging, histopathological and follow-up data, reported in accordance with the CARE guidelines. The patient presented with a right cervicothoracic mass (64 × 45 × 115 mm) circumferentially encasing the common carotid, subclavian and innominate vessels with mediastinal extension. Three tissue procedures were non-diagnostic and FDG PET/CT showed intense uptake (SUVmax 8.15). Radical en bloc resection was undertaken because malignancy could not be excluded. The course was complicated by bilateral vocal-cord paralysis, Horner syndrome, mediastinitis, osteomyelitis and a tracheo-innominate fistula. Histology showed obliterative phlebitis, storiform fibrosis and 74 IgG4-positive plasma cells/HPF (IgG4:IgG 44
Sensorineural hearing loss represents a rare but clinically relevant complication of acute otitis media (AOM). Although intratympanic glucocorticoid (GC) therapy is established in other cochleovestibular disorders, its efficacy in this context remains insufficiently investigated. The aim of this study was to retrospectively assess whether local intratympanic glucocorticoid application is associated with improved clinical outcomes in patients with acute otitis media–related sensorineural hearing loss. A retrospective analysis was performed on patients diagnosed with AOM-associated inner ear involvement between January 2014 and November 2024. Patients were stratified according to GC treatment regimen: (1) intratympanic application of glucocorticoids (IC), (2) no intratympanic application of glucocorticoids (NIC). Subgroup analyses were performed according to the surgical drainage method and severity of inner ear involvement. Primary endpoint was improvement in bone-conduction pure-tone average (BCPTA4). A total of 137 patients were included (IC: 44, NIC: 93). Sex, affected side and subjective improvement did not differ significantly between both groups (all p > 0.08). Both groups demonstrated significant hearing improvement following treatment (p < 0.001). Although numerical trends favored the NIC group across different severity levels, no statistically significant differences in BCPTA4 improvement were observed between treatment modalities (p = 0.8); however, the non-randomized design and baseline group differences limit interpretation. In exploratory analyses, Ventilation Tube Insertion (VT) and CO₂-laser myringotomy (LP) were associated with significant hearing recovery (p < 0.001). Both treatment regimens (IC, NIC) led to significant hearing improvement in AOM-associated inner ear involvement. In this non-randomized comparison, no statistically significant difference in BCPTA4 improvement was detected between groups; however, selection bias, baseline imbalances and residual confounding may have obscured small treatment effects. These findings do not constitute evidence of equivalence.
