Abstract：
OBJECTIVE: to develop a deep learning radiomics graph network (DLRN) that integrates deep learning features extracted from gray scale ultrasonography, radiomics features and clinical features, for distinguishing parotid pleomorphic adenoma (PA) from adenolymphoma (AL) MATERIALS AND METHODS: A total of 287 patients (162 in training cohort, 70 in internal validation cohort and 55 in external validation cohort) from two centers with histologically confirmed PA or AL were enrolled. Deep transfer learning features and radiomics features extracted from gray scale ultrasound images were input to machine learning classifiers including logistic regression (LR), support vector machines (SVM), KNN, RandomForest (RF), ExtraTrees, XGBoost, LightGBM, and MLP to construct deep transfer learning radiomics (DTL) models and Rad models respectively. Deep learning radiomics (DLR) models were constructed by integrating the two features and DLR signatures were generated. Clinical features were further combined with the signatures to develop a DLRN model. The performance of these models was evaluated using receiver operating characteristic (ROC) curve analysis, calibration, decision curve analysis (DCA), and the Hosmer-Lemeshow test.
RESULTS: In the internal validation cohort and external validation cohort, comparing to Clinic (AUC=0.767 and 0.777), Rad (AUC=0.841 and 0.748), DTL (AUC=0.740 and 0.825) and DLR (AUC=0.863 and 0.859), the DLRN model showed greatest discriminatory ability (AUC=0.908 and 0.908) showed optimal discriminatory ability.
CONCLUSIONS: The DLRN model built based on gray scale ultrasonography significantly improved the diagnostic performance for benign salivary gland tumors. It can provide clinicians with a non-invasive and accurate diagnostic approach, which holds important clinical significance and value. Ensemble of multiple models helped alleviate overfitting on the small dataset compared to using Resnet50 alone.
RESULTS: In the internal validation cohort and external validation cohort, comparing to Clinic (AUC=0.767 and 0.777), Rad (AUC=0.841 and 0.748), DTL (AUC=0.740 and 0.825) and DLR (AUC=0.863 and 0.859), the DLRN model showed greatest discriminatory ability (AUC=0.908 and 0.908) showed optimal discriminatory ability.
CONCLUSIONS: The DLRN model built based on gray scale ultrasonography significantly improved the diagnostic performance for benign salivary gland tumors. It can provide clinicians with a non-invasive and accurate diagnostic approach, which holds important clinical significance and value. Ensemble of multiple models helped alleviate overfitting on the small dataset compared to using Resnet50 alone.
摘要：
目的：开发一种深度学习影像组学图形网络（DLRN），该网络集成了从灰度超声检查中提取的深度学习特征，影像组学特征和临床特征，区分腮腺多形性腺瘤(PA)和腺淋巴瘤(AL)材料和方法：共287例患者(162例训练组，纳入了来自两个组织学证实为PA或AL的中心的内部验证队列中的70名和外部验证队列中的55名)。将从灰度超声图像中提取的深度迁移学习特征和影像组学特征输入到机器学习分类器，包括逻辑回归（LR），支持向量机(SVM)，KNN,RandomForest(RF),ExtraTrees,XGBoost,LightGBM,和MLP分别构建深度迁移学习影像组学(DTL)模型和Rad模型。通过整合这两个特征构建深度学习影像组学（DLR）模型，并生成DLR签名。将临床特征与标记进一步组合以开发DLRN模型。使用接收器工作特性(ROC)曲线分析评估了这些模型的性能，校准,决策曲线分析(DCA)，还有Hosmer-Lemeshow测试.
结果：在内部验证队列和外部验证队列中，与诊所相比（AUC=0.767和0.777），拉德（AUC=0.841和0.748），DTL（AUC=0.740和0.825）和DLR（AUC=0.863和0.859），DLRN模型显示出最大的判别能力（AUC=0.908和0.908），显示出最佳的判别能力。
结论：基于灰阶超声建立的DLRN模型显著提高了涎腺良性肿瘤的诊断效能。它可以为临床医生提供一种无创和准确的诊断方法，具有重要的临床意义和价值。与单独使用Resnet50相比，多个模型的集合有助于缓解小数据集上的过度拟合。
结果：在内部验证队列和外部验证队列中，与诊所相比（AUC=0.767和0.777），拉德（AUC=0.841和0.748），DTL（AUC=0.740和0.825）和DLR（AUC=0.863和0.859），DLRN模型显示出最大的判别能力（AUC=0.908和0.908），显示出最佳的判别能力。
结论：基于灰阶超声建立的DLRN模型显著提高了涎腺良性肿瘤的诊断效能。它可以为临床医生提供一种无创和准确的诊断方法，具有重要的临床意义和价值。与单独使用Resnet50相比，多个模型的集合有助于缓解小数据集上的过度拟合。
