基于CT影像组学鉴别骨肌系统良恶性病变的辅助诊断系统  

CT Image of Auxiliary Diagnosis System for Differentiating Benign and Malignant Lesions of Musculoskeletal System Based on Radiomics

作　　者：周平 程丰飞 任波[2] 王健[2] 张冉 柴象飞 ZHOU Ping;CHENG Fengfei;REN Bo;WANG Jian;ZHANG Ran;CHAI Xiangfei(Department of Imaging,Lanling County People’s Hospital,Linyi Shandong 277799,China;Department of Imaging,The First Affiliated Hospital of Xinjiang Medical University,Urumqi Xinjiang 830011,China;Huiying Medical Technology(Beijing)Co.,Ltd.,Beijing 100192,China)

机构地区：[1]山东省兰陵县人民医院影像科,山东临沂277799 [2]新疆医科大学第一附属医院影像科,新疆乌鲁木齐830011 [3]慧影医疗科技(北京)有限公司,北京100192

出　　处：《中国医疗设备》2021年第8期40-44,62,共6页China Medical Devices

基　　金：国家自然科学基金(面上项目)(81974263)。

摘　　要：目的研究一套基于CT影像组学鉴别骨肌系统良恶性病变的辅助诊断系统。方法运用影像组学和医学结合的新兴智能医疗方法研究更精准的智能辅助诊断模型,利用云技术开发更便捷的智能医疗系统,服务于骨肌系统良恶性病变的鉴别和预测。结果在骨肌系统良恶性病变的标准数据库上运用影像组学的方法实现了对影像的自动鉴别,良性和恶性的AUC值分别可达到0.97和0.92的高精度,并充分利用云技术的可扩展性、可移动性和实时性,以模块化的方式开发了一套可辅助临床诊断的智能系统。结论该系统能在短时间内得到可靠的分析模型和智能诊断结果,为临床诊断带来便利的同时,为建立新型的智能分析方法解决骨肌系统良恶性病变的临床鉴别问题提供了新的方向。Objective To study and realize an auxiliary diagnostic system based on CT radiomics to distinguish benign and malignant lesions of the musculoskeletal system.Methods We have applied the emerging intelligent medical methods of combining radiomics and medicine to develop more accurate intelligent auxiliary diagnosis models,and exploited the cloud technology to develop more convenient intelligent medical systems to serve the clinical diagnosis of benign and malignant lesions of the musculoskeletal system.Results On the standard database of benign and malignant lesions of the musculoskeletal system,we achieved the automatic identification of images using the radiomics,and the AUC values of benign and malignant can reach 0.97 and 0.92 respectively with high accuracy,and developed an intelligent system that assist clinical diagnosis in a modular way by making full use of the scalability,mobility and real-time of cloud technology.Conclusion The system can obtain reliable analytical models and intelligent diagnosis results in a short period of time.While bringing convenience to clinical diagnosis.It also provides a new direction to establish a intelligent new analytical method to solve the clinical identification problem of benign and malignant lesions of the skeletal muscle system.

关 键 词：骨肌系统影像 辅助诊断 影像组学 云计算 区域医疗 

分 类 号：R730.44[医药卫生—肿瘤]