汪子琪1,2 王睿卿2 陈告化3 陈伟峰1,2*
1.赣南医科大学;2.赣南医科大学第一附属医院;3.赣州市第五人民医院
摘要(Abstract):
目的比较ChatGPT与DeepSeek在喉结核诊断与治疗中的应用效果。方法回顾性收集我科及协作单位2021年1月至2024年8 月确诊的10例喉结核病例。将匿名化的病历信息输入统一提示词模板后,分别提交至ChatGPT-5与DeepSeek-V3.2。由两名具有副主任医师及以上职称的耳鼻咽喉头颈外科专家,使用Likert量表独立评价模型输出的诊断准确性(3分制)、鉴别诊断完整性(5分制)及治疗建议可信度(5分制),并比较两模型得分差异。结果各评分数据均不符合正态分布。ChatGPT-5在诊断准确性方面显著优于DeepSeek-V3.2[2.50(1.50,3.50)分vs.1.75(1.00,2.50)分,P=0.01]。两者在鉴别诊断完整性方面差异无统计学意义[2.50(2.00,3.00)分vs.2.50(2.38,3.38)分,P=0.31]。ChatGPT-5在治疗建议可信度上略高于DeepSeek-V3.2[3.50(2.50,4.00)分vs.3.00(2.88,3.50)分, P=0.22],但差异未达显著性。结论ChatGPT-5较DeepSeek-V3更适用于喉结核诊治的辅助。两模型均具备良好的临床辅助潜力,但在不同任务类型及处理方法上存在显著差异。
关键词(KeyWords):
大型语言模型;喉结核;人工智能;诊断辅助
参考文献(References):
[1]艾誉峰,刘红兵,徐红,等.原发性和继发性喉结核的临床特征比较分析[J].临床耳鼻咽喉头颈外科杂志,2021,35 (01):38-41.
[2]Andrea M ,Tommaso M ,Anna B , et al.Laryngeal tubercolosis: a case report with focus on voice assessment and review of the literature. [J].Acta otorhinolaryngologica Italica : organo ufficiale della Societa italiana di otorinolaringologia e chirurgia cervico-facciale,2022,42(5):407-414.
[3]Mondillo G,Colosimo S,Perrotta A,et al. Comparative Evaluation of Advanced AI Reasoning Models in Pediatric Clinical Decision Support: ChatGPT O1 vs. DeepSeek-R1[J]. medRxiv, 2025:2021-2025.
[4]Chen M ,Decary M .Artificial intelligence in healthcare: An essential guide for health leaders[J].Healthcare Management Forum,2020,33(1):10-18.
[5]Ningjun Z , Yi Z , Keyong L .Rigid laryngoscope manifestations of 61 cases of modern laryngeal tuberculosis. [ J ] .Experimental and therapeutic medicine , 2017 , 14 (5):5093-5096.
[6]Keyvan K ,Reza M R H .Laryngeal tuberculosis without pulmonary involvement.[J].Caspian journal of internal medicine, 2012,3(1):397-9.
[7]Kimberly W ,Christian F ,Karthik R .Answering head and neck cancer questions: An assessment of ChatGPT responses [J].American Journal of Otolaryngology--Head and Neck Medicine and Surgery,2024,45(1):104085-104085.
[8]Oğuz K ,Erim A P ,Nilda S S ,et al.Is ChatGPT accurate and reliable in answering questions regarding head and neck cancer? [J].Frontiers in Oncology,2023,131256459-1256459.
[9]J R D ,Oluwatobiloba A ,E M L , et al.Evaluation of Oropharyngeal Cancer Information from Revolutionary Artificial Intelligence Chatbot. [ J ] .The Laryngoscope , 2023 , 134 (5):2252-2257.
[10]潘鑫,郜飞,陈亭亭,等.颈部单中心型 Castleman 病临床辅助诊疗中 ChatGPT o1 和 Claude 3.5 Sonnet 的应用比较研究[J].中国耳鼻咽喉颅底外科杂志,2025,31(06):76-82.
[11]黎超,陈优美,段亚妮,等.生成式人工智能在生成影像学报告方面的表现评估[J].新医学,2024,55(11):853-860.
[12]Jian Z ,Ying T ,Xuejun J , et al.Appearance and morphologic features of laryngeal tuberculosis using laryngoscopy: A retrospective cross-sectional study. [J].Medicine , 2020 , 99 (51):e23770-e23770.
[13]Lechien JR,Chiesa-Estomba CM,Baudouin R,Hans S. Accuracy of ChatGPT in head and neck oncological board decisions: preliminary findings. [J].Eur Arch Otorhinolaryngol. 2024;281(4):2105-2114.
[14]邢倩,何达.医疗大语言模型的评价现状及思考[J]. 健康发展与政策研究,2025,28(01):65-72+79.
[15]Lorenzi A,Pugliese G,Maniaci A,et al. Reliability of large language models for advanced head and neck malignancies management: a comparison between ChatGPT 4 and Gemini Advanced[. J].Eur Arch Otorhinolaryngol. 2024;281(9):5001-5006.
[16]Lee JC,Hamill CS,Shnayder Y,Buczek E,Kakarala K,Bur AM. Exploring the Role of Artificial Intelligence Chatbots in Preoperative Counseling for Head and Neck Cancer Surgery. [J].Laryngoscope. 2024;134(6):2757-2761.