摘要
为解决临床上无法在术前预测泌尿系结石类型的问题,提出一种基于患者CT影像来术前预测结石类型的方法,并开发了基于改进残差网络的泌尿系结石类型术前预测辅助系统,实现了结石类型的术前预测。具体工作包括:首先,以Resnet34作为基础网络,改进了池化层、残差块结构和损失函数,并加入了密集连接结构,从而解决了结石误识别的问题;其次,设计了一种双分支多头自注意力模块,增加了结石区域的特征权重,大大提高了模型性能;最后,通过回顾式研究,建立了包含5709张结石CT影像数据集,按照3∶1∶1的比例随机分为训练集、验证集和测试集,用以对模型进行训练和性能验证。经实验验证,提出的改进残差网络在测试集上准确度达到了72.90%,相对于原网络准确度提升了6.38%,F1分数提高了10%。结果表明,该模型能够有效提高泌尿系结石的术前预测精度并具有潜在的临床应用价值。
To solve the problem of clinical inability to predict urinary stone types preoperatively,we propose a method for preoperative prediction of stone types based on CT images,and develop a preoperative prediction aid system for urinary stone types based on improved residual network.This enables preoperative prediction of stone types.The main tasks includes:firstly,Resnet34 is used as the base network with improved pooling layer,residual block structure and loss function,and a dense connection structure is added.Thus,the problem of stone misidentification is solved.Secondly,a two-branch multi-headed self-attentive module was designed to increase the feature weights of the stone region,which greatly improved the model performance.Finally,through a retrospective study,a dataset containing 5709 CT images of stones was created and randomly divided into a training set,a validation set and a test set in the ratio of 3∶1∶1,which was used for training and performance validation of the model.The proposed improved residual network was experimentally verified to be 72.90%accurate on the test set,with a 6.38%improvement in accuracy and a 10%increase in F1 score.The results showed that the model can effectively improve the accuracy of preoperative prediction of urinary stones and has potential clinical application.
作者
刘琨
王向辉
崔振宇
杨昆
Liu Kun;Wang Xianghui;Cui Zhenyu;Yang Kun(College of Quality and Technical Supervision,Hebei University,Baoding 071002,China;Department of Urology,Affiliated Hospital of Hebei University,Baoding 071000,China;Hebei Technology Innovation Center for Lightweight of New Energy Vehicle Power System,Baoding 071002,China;Postdoctoral Research Station of Optical Engineering,Hebei University,Baoding 071002,China)
出处
《电子测量技术》
北大核心
2023年第18期147-154,共8页
Electronic Measurement Technology
基金
河北省高层次人才资助项目(B20190030010)
河北省自然科学基金(H2019201378)
河北大学校长基金(XZJJ201917)
河北大学医学学科培育项目(2021X07)资助
关键词
泌尿系结石
术前预测
残差网络
图像分类
urological stone
preoperative prediction
residual network
image classification
作者简介
刘琨,博士,讲师,硕士生导师,主要研究方向为人工智能医学图像处理。E-mail:liukun15166@hbu.edu.cn;王向辉,硕士研究生,主要研究方向为人工智能医学图像处理。E-mail:1965452981@qq.com;崔振宇,博士研究生,副主任,主要研究方向为泌尿系结石、肿瘤、前列腺疾病的诊治。E-mail:cuizhenyu615@163.com;通信作者:杨昆,博士,教授,博士生导师,主要研究方向为人工智能医学图像处理。E-mail:hbu0126@163.com
