伍晓妮

甲状腺结节的超声评估已经形成较成熟的风险分层体系，但实际判读仍受到切面选择、图像质量和操作者经验影响。人工智能进入甲状腺超声后，应用重点已由单幅图像分类逐步转向征象识别、TI-RADS分级、穿刺选择及颈部淋巴结评估。现有文献表明，多中心训练、多视图分析和多模态信息整合能够补充传统超声判读，不过模型在不同设备、特殊病理类型和动态扫查场景中的稳定性仍需验证。本文结合甲状腺结节的实际诊疗环节，对AI在图像识别、风险分层、穿刺决策和术前评估中的应用进行梳理，并讨论模型解释、外部验证和临床部署中的关键问题。现阶段，AI更适合作为超声医师的辅助工具，其结果仍需与实时扫查、病理结果及患者个体风险共同分析。

甲状腺癌；人工智能；超声；深度学习；TI-RADS；风险分层

[1] Tessler FN,Middleton WD,Grant EG,et al.ACR Thyroid Imaging,Reporting and Data System(TI-RADS):White Paper of the ACR TI-RADS Committee.J Am Coll Radiol,2017,14(5):587-595. [2] Chen C,Jiang Y,Yao J,et al.Deep learning to assist composition classification and thyroid solid nodule diagnosis:a multicenter diagnostic study.Eur Radiol,2024,34(4):2323-2333. [3] Vahdati S,Khosravi B,Robinson KA,et al.A multi-view deep learning model for thyroid nodules detection and characterization in ultrasound imaging.Bioengineering,2024,11(7):648. [4] Zhou T,Xu L,Shi J,et al.US of thyroid nodules:can AI-assisted diagnostic system compete with fine needle aspiration?Eur Radiol,2024,34(2):1324-1333. [5] Wang X,Qi Y,Zhang X,et al.Ultrasound-based artificial intelligence for predicting cervical lymph node metastasis in papillary thyroid cancer:a systematic review and meta-analysis.Front Endocrinol,2025,16:1570811. [6] Li Y,Li X,Luo Y,et al.Multiparametric ultrasound techniques are superior to AI-assisted ultrasound for assessment of solid thyroid nodules:a prospective study.Endocrine,2025,90(1):225-236. [7] Athreya S,Melehy A,Suthahar SSA,et al.Combining ultrasound imaging and molecular testing in a multimodal deep learning model for risk stratification of indeterminate thyroid nodules.Thyroid,2025,35(5):590-594.

点击下载PDF