Deep learning-assisted diagnosis of benign and malignant parotid tumors based on contrast-enhanced CT: a multicenter study

作者全名："Yu, Qiang; Ning, Youquan; Wang, Anran; Li, Shuang; Gu, Jinming; Li, Quanjiang; Chen, Xinwei; Lv, Fajin; Zhang, Xiaodi; Yue, Qiang; Peng, Juan"

作者地址："[Yu, Qiang; Ning, Youquan; Wang, Anran; Gu, Jinming; Li, Quanjiang; Chen, Xinwei; Lv, Fajin; Peng, Juan] Chongqing Med Univ, Affiliated Hosp 1, Dept Radiol, 1 Youyi Rd, Chongqing 400016, Peoples R China; [Li, Shuang; Yue, Qiang] Sichuan Univ, West China Hosp, Dept Radiol, 37 Guoxue Lane, Chengdu 610041, Peoples R China; [Zhang, Xiaodi] Philips Healthcare, Chengdu 610041, Peoples R China"

通信作者："Peng, J (通讯作者)，Chongqing Med Univ, Affiliated Hosp 1, Dept Radiol, 1 Youyi Rd, Chongqing 400016, Peoples R China.; Yue, Q (通讯作者)，Sichuan Univ, West China Hosp, Dept Radiol, 37 Guoxue Lane, Chengdu 610041, Peoples R China."

来源：EUROPEAN RADIOLOGY

ESI学科分类：CLINICAL MEDICINE

WOS号：WOS:000969367700001

JCR分区：Q1

影响因子：4.7

年份：2023

卷号：　

期号：　

开始页：　

结束页：　

文献类型：Article; Early Access

关键词：Deep learning; Machine learning; Radiomics; Parotid neoplasms; Computed tomography

摘要："ObjectivesTo develop deep learning-assisted diagnosis models based on CT images to facilitate radiologists in differentiating benign and malignant parotid tumors.MethodsData from 573 patients with histopathologically confirmed parotid tumors from center 1 (training set: n = 269; internal-testing set: n = 116) and center 2 (external-testing set: n = 188) were retrospectively collected. Six deep learning models (MobileNet V3, ShuffleNet V2, Inception V3, DenseNet 121, ResNet 50, and VGG 19) based on arterial-phase CT images, and a baseline support vector machine (SVM) model integrating clinical-radiological features with handcrafted radiomics signatures were constructed. The performance of senior and junior radiologists with and without optimal model assistance was compared. The net reclassification index (NRI) and integrated discrimination improvement (IDI) were calculated to evaluate the clinical benefit of using the optimal model.ResultsMobileNet V3 had the best predictive performance, with sensitivity increases of 0.111 and 0.207 (p < 0.05) in the internal- and external-testing sets, respectively, relative to the SVM model. Clinical benefit and overall efficiency of junior radiologist were significantly improved with model assistance; for the internal- and external-testing sets, respectively, the AUCs improved by 0.128 and 0.102 (p < 0.05), the sensitivity improved by 0.194 and 0.120 (p < 0.05), the NRIs were 0.257 and 0.205 (p < 0.001), and the IDIs were 0.316 and 0.252 (p < 0.001).ConclusionsThe developed deep learning models can assist radiologists in achieving higher diagnostic performance and hopefully provide more valuable information for clinical decision-making in patients with parotid tumors."

基金机构："Foundation of Science and Technology Bureau of Yuzhong District, Chongqing, China [20190111]; Natural Science Foundation of Chongqing, China [cstc2021jcyj-msxmX0020]"

基金资助正文："AcknowledgementsThis study received support from the Foundation of Science and Technology Bureau of Yuzhong District, Chongqing, China (Grant No. 20190111) and the Natural Science Foundation of Chongqing, China (Grant No. cstc2021jcyj-msxmX0020). We thank the American Journal Experts (AJE) for their assistance with language editing, and we appreciate the OnekeyAI platform and its developers, as well as all of the individuals who participated in these studies and each of the researchers and technicians who made this work possible."