基于甲状腺超声图像建立甲状腺乳头状癌中央区淋巴结转移人工智能诊断模型

目的 基于甲状腺超声图像建立甲状腺乳头状癌中央区淋巴结转移人工智能诊断模型。方法 回顾性分析2018年1至12月于中国人民解放军总医院第一医学中心行甲状腺切除及颈部中央区淋巴结清扫的309例甲状腺乳头状癌(PTC)患者的临床资料及超声图像,病理结果为金标准。所有病例被分为训练集(265例)、测试集(44例)。基于深度学习方法建立甲状腺超声图像预测PTC患者中央区淋巴结转移的计算机辅助诊断系统。在测试集中评估该系统的诊断性能。结果 在测试集中,本模型预测PTC患者中央区淋巴结转移的准确性、敏感性、特异性和受试者工作特征曲线下面积可达80%、76%、83%、0.794。结论 基于深度学习的人工智能诊断模型可用于诊断甲状腺乳头状癌患者中央区淋巴结转移,可为临床选择治疗方案提供依据。

A Thyroid Ultrasound Image-based Artificial Intelligence Model for Diagnosis of Central Compartment Lymph Node Metastasis in Papillary Thyroid Carcinoma

Objective To establish an artificial intelligence model based on B-mode thyroid ultrasound images to predict central compartment lymph node metastasis(CLNM)in patients with papillary thyroid carcinoma(PTC). Methods We retrieved the clinical manifestations and ultrasound images of the tumors in 309 patients with surgical histologically confirmed PTC and treated in the First Medical Center of PLA General Hospital from January to December in 2018.The datasets were split into the training set and the test set.We established a deep learning-based computer-aided model for the diagnosis of CLNM in patients with PTC and then evaluated the diagnosis performance of this model with the test set. Result The accuracy,sensitivity,specificity,and area under receiver operating characteristic curve of our model for predicting CLNM were 80%,76%,83%,and 0.794,respectively. Conclusion Deep learning-based radiomics can be applied in predicting CLNM in patients with PTC and provide a basis for therapeutic regimen selection in clinical practice.