基于DenseNet_ECA的儿童异常胸片筛查  

Screening Abnormal Chest X-Ray in Children Based on DenseNet_ECA

作　　者：段梦宇 吴英飞[1,2] 袁贞明 俞刚[3,4] DUAN Meng-yu;WU Ying-fei;UAN Zhen-ming;YU Cang(School of Information Science and Technology,Hangzhou Normal University,Hangzhou Zhejiang 311121,China;Engineering Research Center of Mobile Health Management System,Ministry of Education,Hangzhou Zhejiang 311121,China;Department of Data and Information,Children's Hospital,Zhejiang University School of Medicine,Hangzhou Zhejiang 310052,China;National Clinical Research Center for Child Health,Hangzhou Zhejiang 310052,China)

机构地区：[1]杭州师范大学信息科学与技术学院,浙江杭州311121 [2]移动健康管理系统教育部工程研究中心,浙江杭州311121 [3]浙江大学医学院附属儿童医院数据信息部,浙江杭州310052 [4]国家儿童健康与疾病临床医学研究中心,浙江杭州310052

出　　处：《计算机仿真》2024年第1期261-265,343,共6页Computer Simulation

基　　金：国家重点研发计划(.2019YFE0126200);国家自然科学基金面上项目(62076218)。

摘　　要：胸片是筛查儿童肺部异常最常见、最容易获得的低成本成像方式。然而,在部分医疗资源匮乏地区,由于有经验的放射科医生数量稀少,导致胸片的解读效率低下,易造成对肺部异常患儿的漏诊、误诊。因此以儿童健康及异常胸片为研究对象,通过使用ECA注意力机制及PReLU激活函数对DenseNet进行改进,提出一种用于儿童异常胸片筛查任务的DenseNet_ECA模型。实验结果表明,上述模型对于儿童健康、异常胸片的分类效果优于常用卷积神经网络模型,分类准确率、灵敏度、特异性分别可达93.57%,91.47%,95.83%,参数量仅为6.96M。上述模型能够帮助医生进行儿童异常胸片的预先筛查,可有效降低临床阅片压力,提高医生诊断效率。Chest X-ray is the most common and easily available low-cost imaging method for screening pulmonary abnormalities in children.However,in some areas where medical resources are scarce,due to the small number of experienced radiologists,the interpretation of chest X-ray is inefficient,which is easy to cause missed diagnosis and misdiagnosis of children with pulmonary abnormalities.Therefore,this paper took children's health and abnormal chest X-ray as the research object,improved DenseNet by using ECA attention mechanism and PReLU activation function,and proposed a deep learning modelDenseNet_ECA for children's abnormal chest X-ray screening task.The experimental results show that the classification effect of this model for children's healthy and abnormal chest Xrayis better than that of common convolutional neural networkmodels.The classification accuracy,sensitivity and specificity can reach 93.57%,91.47%and 95.83%respectively,and the parameter quantity is only 6.96M.The model can help doctors to pre-screen abnormalchest X-ray of children,effectively reduce the pressure ofclinical film reading and improve the diagnostic efficiency of doctors.

关 键 词：儿童异常胸片筛查 注意力机制 仿真 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]