Wang Xiheng, Sun Zhaoyong, Xue Huadan, Qu Taiping, Cheng Sihang, Li Juan, Li Yatong, Mao Li, Li Xiuli, Zhu Liang, Li Xiao, Zhang Longjing, Jin Zhengyu, Yu Yizhou
Department of Radiology, Peking Union Medical College Hospital, Peking Union Medical College, Chinese Academy of Medical Sciences, Shuaifuyuan No.1, Wangfujing Street, Dongcheng District, Beijing, 100730, People's Republic of China.
Deepwise AI Lab, Deepwise Inc., Beijing, 100080, People's Republic of China.
Abdom Radiol (NY). 2022 Jun;47(6):2135-2147. doi: 10.1007/s00261-022-03479-4. Epub 2022 Mar 27.
To develop a deep learning model (DLM) to improve readers' interpretation and speed in the differentiation of pancreatic cystic lesions (PCLs) on dual-phase enhanced CT, and a low contrast media dose, external testing set validated the model.
Dual-phase enhanced CT images of 363 patients with 368 PCLs obtained from two centers were retrospectively assessed. Based on the examination date, a training and validation set of 266 PCLs, an internal testing set of 52 PCLs were designated from center 1. An external testing set included 50 PCLs from center 2. Clinical and radiological characteristics were compared. The DLM was developed using 3D specially designed densely connected convolutional networks for PCL differentiation. Radiomic features were extracted to build a traditional radiomics model (RM). Performance of the DLM, traditional RM, and three readers was compared.
The accuracy for differential diagnosis was 0.904 with DLM, which was the highest in the internal testing set. Accuracy differences between the DLM and senior radiologist were not significant both in the internal and external testing set (both p > 0.05). With the help of the DLM, the accuracy and specificity of the junior radiologist were significantly improved (all p < 0.05), and all readers' diagnostic time was shortened (all p < 0.05).
The DLM achieved senior radiologist-level performance in differentiating benign and malignant PCLs which could improve the junior radiologist's interpretation and speed of PCLs on CT.
开发一种深度学习模型（DLM），以提高读者在双期增强CT上鉴别胰腺囊性病变（PCL）的准确性和速度，并使用低对比剂剂量的外部测试集对该模型进行验证。
回顾性分析来自两个中心的363例患者的368个PCL的双期增强CT图像。根据检查日期，从中心1指定了266个PCL的训练和验证集，52个PCL的内部测试集。外部测试集包括来自中心2的50个PCL。比较临床和放射学特征。使用专门设计的用于PCL鉴别的3D密集连接卷积网络开发DLM。提取放射组学特征以建立传统放射组学模型（RM）。比较了DLM、传统RM和三位读者的表现。
DLM在内部测试集中鉴别诊断的准确率为0.904，是最高的。DLM与高级放射科医生在内部和外部测试集中的准确率差异均无统计学意义（均p>0.05）。在DLM的帮助下，初级放射科医生的准确率和特异性均显著提高（均p<0.05），所有读者的诊断时间均缩短（均p<0.05）。
DLM在鉴别PCL的良恶性方面达到了高级放射科医生的水平，可提高初级放射科医生对CT上PCL的解读能力和速度。
