基于超声影像的下腔静脉自动识别与定位  

Automated identification and localization of inferior vena cava based on ultrasound images

作　　者：杨敬涵 陈子叶 孙景源 曹文[4] 吕朝阳 李硕[4] 李明秋[4] 张普 徐径舟 周昌 杨宇祥 张甫 李庆利[1] 郭瑞君[4] 陈建刚 YANG Jinghan;CHEN Ziye;SUN Jingyuan;CAO Wen;LÜChaoyang;LI Shuo;LI Mingqiu;ZHANG Pu;XU Jingzhou;ZHOU Chang;YANG Yuxiang;ZHANG Fu;LI Qingli;GUO Ruijun;CHEN Jiangang(Shanghai Key Laboratory of Multidimensional Information Processing,East China Normal University,Shanghai 200241,China;School of Advanced Technology,Xi’an Jiaotong-Liverpool University,Suzhou 215123,Jiangsu,China;Department of Computer Science,University of Liverpool,Liverpool L693BX,England;Department of Ultrasound,Beijing Chao-yang Hospital,Capital Medical University,Beijing 100020,China;Department of General Surgery,Baoding Orthopaedics Hospital,Baoding 071000,Hebei,China;Department of Electronic Engineering,School of Engineering and Design,Hunan Normal University,Changsha 410081,Hunan,China)

机构地区：[1]华东师范大学上海市多维度信息处理重点实验室,上海200241 [2]西交利物浦大学智能工程学院,苏州215123 [3]利物浦大学计算机科学系,利物浦L693BX [4]首都医科大学附属北京朝阳医院超声医学科,北京100020 [5]保定骨科医院普外科,保定071000 [6]湖南师范大学工程与设计学院电子工程系,长沙410081

出　　处：《海军军医大学学报》2024年第9期1107-1112,共6页Academic Journal of Naval Medical University

基　　金：国家自然科学基金(82151318,32201134,32171366,31671002);上海市“科技创新行动计划”医学创新研究专项(21Y11902500);2023年地方科学发展基金(XZ202301YD0032C);吉林省科技发展计划项目(20230204094YY)。

摘　　要：目的探究基于下腔静脉(IVC)超声影像的IVC自动识别与直径测量方法。方法提出一种基于拓扑结构与自动跟踪算法的IVC自动识别与定位方法,采用跟踪方式对数据集进行IVC识别和连续定位,以提高测量的效率和准确性。以超声医师的人工测量结果作为金标准,在采集自ICU的18例患者的18组超声数据集上进行测试。结果自动化算法识别准确率为94.44%(17/18),IVC直径自动测量误差在±1.96s(s为标准差)范围内,表明自动测量方法可替代人工测量方法。结论本研究提出的基于拓扑结构与自动跟踪算法的IVC自动识别与定位算法具有较高的识别成功率及IVC直径测量准确性,可以辅助临床医师进行IVC的识别定位与直径测量。Objective To explore the automated identification and diameter measurement methods for inferior vena cava(IVC)based on clinical ultrasound images of IVC.Methods An automated identification and localization method based on topology and automatic tracking algorithm was proposed.Tracking algorithm was used for identifying and continuously locating to improve the efficiency and accuracy of measurement.Tests were conducted on 18 sets of ultrasound data collected from 18 patients in intensive care unit(ICU),with clinicians’measurements as the gold standard.Results The recognition accuracy of the automated method was 94.44%(17/18),and the measurement error of IVC diameter was within the range of±1.96s(s was the standard deviation).The automated method could replace the manual method.Conclusion The proposed IVC automated identification and localization algorithm based on topology and automatic tracking algorithm has high recognition success rate and IVC diameter measurement accuracy.It can assist clinicians in identifying and locating IVC,so as to improve the accuracy of IVC measurement.

关 键 词：超声图像 下腔静脉 血管识别与定位 自动化测量 

分 类 号：R445.1[医药卫生—影像医学与核医学]