基于一维块匹配算法的气胸超声诊断算法  

Pneumothorax Diagnosis Algorithm Based on One-dimensional Block Matching Algorithm

作　　者：於天祥 马一博[2] 闫士举[1] 方明娟 邓爽 庄灏喆 周婷婷 Yu Tianxiang;Ma Yibo;Yan Shiju;Fang Mingjuan;Deng Shuang;Zhuang Haozhe;Zhou Tingting(School of Health Science and Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;Department of Ultrasound,The Third Affiliated Hospital of Soochow University,Changzhou,Jiangsu 213003,China)

机构地区：[1]上海理工大学健康科学与工程学院,上海市200093 [2]苏州大学附属第三医院超声医学科,江苏省常州市213003

出　　处：《中国超声医学杂志》2022年第12期1362-1364,共3页Chinese Journal of Ultrasound in Medicine

摘　　要：目的 提出一种基于一维块匹配算法的气胸超声诊断算法以辅助医务人员诊断气胸。方法 使用由合作医院提供的669段肺部超声视频测试算法性能并由合作医师对这些视频进行分类。本文利用一维块匹配算法提取视频特征,并使用K最近邻算法(KNN)进行分类。结果 在对数据预处理的情况下,混合数据(征象典型+征象非典型)的分类正确率为94.10%,三类视频平均灵敏度为94.10%,特异度为97.03%,曲线下面积(AUC)为0.96。结论 本文提出的算法在分类视频并诊断气胸方面具有较好的性能。Objective To develop a diagnosis algorithm based on the one-dimensional block matching algorithm which can be used to aid doctors in pneumothrax diagnosis. Methods A database including 669 lung ultrasonography videos provided by the cooperative hospital was used to test and the cooperative doctor manually divided the videos. One-dimensional block matching algorithm was used to analyse characters and K-Nearest Neighbor was used to classify videos. Results With preprocessing the data, the accuracy, true positive rate, true negative rate and AUC were 94.10%,94.10%,97.03% and 0.96. Conclusions The algorithm developed by the study is reliable in the classification of videos and diagnosis of pneumothorax.

关 键 词：肺部超声 气胸诊断 块匹配算法 运动估计 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] R445.1[自动化与计算机技术—计算机科学与技术]