Oral diseases · 2025
Accuracy of ChatGPT-4 Plus in Providing Information on Oral Cancer Management.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Yilmaz Busra, Kahraman Emine Nur, Brennan Michael T, Grewal Amardeep S, Aktas Aynur
Аннотация
OBJECTIVE: Artificial intelligence (AI)-driven large language models, such as Chat Generative Pre-Trained Transformer (ChatGPT)-4 Plus, are increasingly used for patient education and clinical decision support in oral oncology, although their accuracy in oral cancer (OC) management remains uncertain. This study evaluates the accuracy of ChatGPT-4 Plus responses to clinically relevant questions regarding OC diagnosis, treatment, recovery, and prevention.
METHODS: A cross-sectional study assessed 65 clinically relevant OC-related questions using a paid ChatGPT-4 Plus subscription without modifications. Three oral medicine specialists and one radiation oncologist rated accuracy on a four-point scoring system. Interrater reliability was measured with the intraclass correlation coefficient (ICC), and chi-square tests were used for comparisons.
RESULTS: Among 65 questions, 63% of responses were Score 1, with none rated as Score 4. Score 1 was most frequent in Recovery (72%), followed by Treatment (62%), Prevention (60%), and Diagnosis (55%). Scores 2 and 3 responses were highest in Diagnosis (45%). Recovery had significantly higher Score 1 responses than Diagnosis (p < 0.05), while other comparisons were not significant. ICC ranged from 0.85 to 0.93.
CONCLUSIONS: ChatGPT-4 Plus provided accurate responses to clinically relevant OC-related questions, particularly regarding recovery. However, diagnostic inconsistencies highlight the need for clinician oversight before integrating AI into practice.
METHODS: A cross-sectional study assessed 65 clinically relevant OC-related questions using a paid ChatGPT-4 Plus subscription without modifications. Three oral medicine specialists and one radiation oncologist rated accuracy on a four-point scoring system. Interrater reliability was measured with the intraclass correlation coefficient (ICC), and chi-square tests were used for comparisons.
RESULTS: Among 65 questions, 63% of responses were Score 1, with none rated as Score 4. Score 1 was most frequent in Recovery (72%), followed by Treatment (62%), Prevention (60%), and Diagnosis (55%). Scores 2 and 3 responses were highest in Diagnosis (45%). Recovery had significantly higher Score 1 responses than Diagnosis (p < 0.05), while other comparisons were not significant. ICC ranged from 0.85 to 0.93.
CONCLUSIONS: ChatGPT-4 Plus provided accurate responses to clinically relevant OC-related questions, particularly regarding recovery. However, diagnostic inconsistencies highlight the need for clinician oversight before integrating AI into practice.