实用肿瘤杂志2025，Vol.40Issue(5)：447-451,5.DOI:10.13267/j.cnki.syzlzz.2025.065
C-TIRADS和人工智能辅助诊断系统在甲状腺结节诊断中的价值
Value of C-TIRADS and artificial intelligence assisted diagnosis system in diagnosis of thyroid nodules
摘要
Abstract
Objective To explore the value of Chinese-Thyroid Imaging Reporting and Data System(C-TIRADS)and artificial intelli-gence(Al)assisted diagnosis system in the identification of benign and malignant thyroid nodules,and analyze the effectiveness of the AI assisted diagnosis system for nodule diagnosis based on different ultrasound features.Methods A total of 1 039 thyroid nodules were col-lected from 977 patients who underwent thyroid surgery in the Second Hospital of Hebei Medical University from May 2020 to May 2023.Using the results of histopathological examination as the gold standard,504 nodules were benign,and 535 nodules were malignant.Re-ceiver operating characteristic(ROC)curves were used to analyze the value of physicians and AI assisted diagnosis system in the diagnosis of thyroid nodules.The thyroid nodules were classified according to the orientation,margin,structure,echo,and focal strong echo.Results The areas under the ROC curves of physicians and AI assisted diagnosis system diagnosing the thyroid nodules were 0.912 and 0.908,re-spectively.The diagnostic efficacy of the AI assisted diagnosis system was comparable to physicians and the diagnostic accuracy was higher for thyroid nodules with hypoechoic feature(88.6%vs 81.9%,P=0.001),ill-defined margin(90.3%vs 84.5%,P=0.023),and macrocalci-fications(85.8%vs 68.1%,P<0.01).Conclusions The AI assisted diagnosis system has a high value for the diagnosis of thyroid nodules,and can make better diagnosis for thyroid nodules with hypoechoic feature,ill-defined margin,and macrocalcifications.关键词
甲状腺结节/超声检查/甲状腺影像报告和数据系统/人工智能Key words
thyroid nodule/ultrasound/Chinese-Thyroid Imaging Reporting and Date System/artificial intelligence引用本文复制引用
张苗苗,高子逸,吴琼,王影..C-TIRADS和人工智能辅助诊断系统在甲状腺结节诊断中的价值[J].实用肿瘤杂志,2025,40(5):447-451,5.基金项目
河北省省级科技计划资助(21377777D) （21377777D）
