深度学习作为甲状腺细针穿刺细胞学辅助诊断工具的应用
本研究用393个甲状腺结节的148,395张细针穿刺细胞学显微图像训练并验证EfficientNetV2-L图像分类模型,评估AI对甲状腺细胞学的分类准确性,并探索其在意义不明确的不典型(AUS)结节中的预测价值。结果显示多数类别PR AUC>0.95,低分化癌最难识别,AUS结节分析敏感度94.7%、特异度14.4%。
要点
- AI对多数甲状腺细胞学类别判别准确,PR AUC>0.95。
- 低分化甲状腺癌识别最差,PR AUC 0.49、召回率35.4%。
- AUS结节AI分析敏感度94.7%、特异度14.4%。
核心结论
- 基于深度学习的甲状腺FNAC图像分析在多数类别中判别准确,有望用于AUS与滤泡性肿瘤结节的临床管理。
- AUS结节分析敏感度高但特异度低,提示其更适合作为辅助而非独立诊断工具。
研究设计与样本
- 回顾性获取393个甲状腺结节的148,395张FNAC显微图像用于训练与验证。
- 采用EfficientNetV2-L作为图像分类模型,并对35个AUS结节进行AI预测。
主要发现
- 除低分化甲状腺癌(PR AUC=0.49)与甲状腺髓样癌(PR AUC=0.91)外,其余类别PR AUC均>0.95。
- 低分化甲状腺癌召回率最低(35.4%),难以与乳头状癌、髓样癌和滤泡状癌区分。
- 滤泡性腺瘤与滤泡状癌的召回率分别为86.7%和93.9%。
- t-SNE二维映射中,淋巴瘤、滤泡性腺瘤和未分化癌分别被分为三组、两组和两组。
- AUS结节分析敏感度94.7%、特异度14.4%、阳性预测值56.3%、阴性预测值66.7%。
关键数值与推荐
- 多数类别PR AUC>0.95,低分化甲状腺癌PR AUC=0.49,髓样癌PR AUC=0.91。
- 低分化甲状腺癌召回率35.4%;滤泡性腺瘤与滤泡状癌召回率分别为86.7%与93.9%。
- AUS结节:敏感度94.7%、特异度14.4%、阳性预测值56.3%、阴性预测值66.7%。
临床怎么用
- AI分析可作为甲状腺FNAC的辅助诊断工具,尤其用于AUS与滤泡性肿瘤结节的临床管理参考。
- 可探索以在线AI平台形式用于甲状腺细胞学会诊。
- 低分化甲状腺癌识别能力有限,需结合常规形态学与其他检查综合判断。
局限性
- 模型对低分化甲状腺癌识别能力低,难以与多种甲状腺癌鉴别。
- AUS结节分析特异度仅14.4%,阳性预测值56.3%,误判风险需关注。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND: Several studies have used artificial intelligence (AI) to analyze cytology images, but AI has yet to be adopted in clinical practice. The objective of this study was to demonstrate the accuracy of AI-based image analysis for thyroid fine-needle aspiration cytology (FNAC) and to propose its application in clinical practice. METHODS: In total, 148,395 microscopic images of FNAC were obtained from 393 thyroid nodules to train and validate the data, and EfficientNetV2-L was used as the image-classification model. The 35 nodules that were classified as atypia of undetermined significance (AUS) were predicted using AI training. RESULTS: The precision-recall area under the curve (PR AUC) was >0.95, except for poorly differentiated thyroid carcinoma (PR AUC = 0.49) and medullary thyroid carcinoma (PR AUC = 0.91). Poorly differentiated thyroid carcinoma had the lowest recall (35.4%) and was difficult to distinguish from papillary thyroid carcinoma, medullary thyroid carcinoma, and follicular thyroid carcinoma. Follicular adenomas and follicular thyroid carcinomas were distinguished from each other by 86.7% and 93.9% recall, respectively. For two-dimensional mapping of the data using t-distributed stochastic neighbor embedding, the lymphomas, follicular adenomas, and anaplastic thyroid carcinomas were divided into three, two, and two groups, respectively. Analysis of the AUS nodules showed 94.7% sensitivity, 14.4% specificity, 56.3% positive predictive value, and 66.7% negative predictive value. CONCLUSIONS: The authors developed an AI-based approach to analyze thyroid FNAC cases encountered in routine practice. This analysis could be useful for the clinical management of AUS and follicular neoplasm nodules (e.g., an online AI platform for thyroid cytology consultations).
引用
- Hirokawa M, Niioka H, Suzuki A, et al. Application of deep learning as an ancillary diagnostic tool for thyroid FNA cytology. Cancer Cytopathol. 2022. PMID: 36524985. doi:10.1002/cncy.22669
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。