2025/09/01 by Mitat Selçuk Bozhöyük, Levent Yücel · 1 voice
Medicine · #Artificial Intelligence in Healthcare and Education #Radiomics and Machine Learning in Medical Imaging #Surgical Simulation and Training
paper · doi:10.1017/s0022215125103368
openalex created_date 2025/09/01 · openalex publication_date 2025/09/01 · openalex updated_date 2026/07/25
Abstract Objectives This study aimed to evaluate the quality of information provided by artificial intelligence (AI) applications regarding ENT surgeries and usability for patients. Methods ChatGPT 4.0, GEMINI 1.5 Flash, Copilot, Claude 3.5 Sonnet and DeepSeek-R1 were asked to provide detailed responses to patient-oriented questions about 15 ENT surgeries. Each AI application was queried three times, with a 3-day interval between each session. Two ENT specialists evaluated all responses using the Quality Analysis of Medical Artificial Intelligence (QAMAI) tool. Results Average QAMAI scores for each AI application were as follows: ChatGPT 4.0 (27.56 ± 1.20), GEMINI 1.5 Flash (26.24 ± 1.26), Copilot (26.84 ± 1.35), Claude 3.5 Sonnet (28.24 ± 0.77) and DeepSeek-R1 (28.13 ± 0.84). A statistically significant difference was found among the applications ( p < 0.001). ICC analysis indicated high stability across evaluations conducted for all five AI applications ( p < 0.001). Conclusion AI has the potential to provide patients with accurate and consistent information about ENT surgeries, yet differences in QAMAI scores show that information quality varies between platforms.