Zhu Wan, Huang Xingzhi, Qi Qi, Wu Zhenghua, Min Xiang, Zhou Aiyun, Xu Pan
Departments of Ultrasonography, The First Affiliated Hospital of Nanchang University, Nanchang, Jiangxi, China.
Departments of Head and Neck Otolaryngology, The First Affiliated Hospital of Nanchang University, Nanchang, Jiangxi, China.
J Oncol. 2022 Jun 17;2022:7133972. doi: 10.1155/2022/7133972. eCollection 2022.
To evaluate the ability of artificial neural network- (ANN-) based ultrasound radiomics to predict large-volume lymph node metastasis (LNM) preoperatively in clinical N0 disease (cN0) papillary thyroid carcinoma (PTC) patients.
From January 2020 to April 2021, 306 cN0 PTC patients admitted to our hospital were retrospectively reviewed and divided into a training ( = 183) cohort and a validation cohort ( = 123) in a 6 : 4 ratio. Radiomic features quantitatively extracted from ultrasound images were pruned to train one ANN-based radiomic model and three conventional machine learning-based classifiers in the training cohort. Furthermore, an integrated model using ANN was constructed for better prediction. Meanwhile, the prediction of the two models was evaluated in the papillary thyroid microcarcinoma (PTMC) and conventional papillary thyroid cancer (CPTC) subgroups.
The radiomic model showed better discrimination than other classifiers for large-volume LNM in the validation cohort, with an area under the receiver operating characteristic curve (AUROC) of 0.856 and an area under the precision-recall curve (AUPR) of 0.381. The performance of the integrated model was better, with an AUROC of 0.910 and an AUPR of 0.463. According to the calibration curve and decision curve analysis, the radiomic and integrated models had good calibration and clinical usefulness. Moreover, the models had good predictive performance in the PTMC and CPTC subgroups.
ANN-based ultrasound radiomics could be a potential tool to predict large-volume LNM preoperatively in cN0 PTC patients.
评估基于人工神经网络（ANN）的超声影像组学在临床N0期疾病（cN0）的乳头状甲状腺癌（PTC）患者中术前预测大体积淋巴结转移（LNM）的能力。
回顾性分析2020年1月至2021年4月我院收治的306例cN0 PTC患者，并按照6∶4的比例分为训练队列（n = 183）和验证队列（n = 123）。从超声图像中定量提取影像组学特征，在训练队列中筛选影像组学特征以训练一个基于ANN的影像组学模型和三个基于传统机器学习的分类器。此外，构建一个使用ANN的集成模型以实现更好的预测。同时，在甲状腺微小癌（PTMC）和传统乳头状甲状腺癌（CPTC）亚组中评估这两个模型的预测能力。
在验证队列中，影像组学模型对大体积LNM的鉴别能力优于其他分类器，其受试者操作特征曲线下面积（AUROC）为0.856, 精确召回率曲线下面积（AUPR）为0.381。集成模型的性能更好，AUROC为0.910，AUPR为0.463。根据校准曲线和决策曲线分析，影像组学模型和集成模型具有良好的校准和临床实用性。此外，这些模型在PTMC和CPTC亚组中具有良好的预测性能。
基于ANN的超声影像组学可能是术前预测cN0 PTC患者大体积LNM的潜在工具。
