摘要
提出一种基于深度残差网络的银屑病分类诊断模型。首先采用数据增强、银屑病图片大小调整和TFRecord编码等技术对网络的输入进行预处理,然后设计了一个34层的深度残差网络(ResNet-34)来对银屑病的特征进行提取。此外,采用交叉熵作为ResNet-34的损失函数来衡量模型的准确性,并利用Adam算法作为优化器来对ResNet-34进行训练,最终得到一个优化的ResNet-34模型用于银屑病诊断。基于K折交叉验证的实验结果表明,所提模型在召回率和ROC曲线方面的性能优于其他诊断方法,可以为银屑病数据分析、疾病预防提供技术支持。
A classification and diagnosis model of psoriasis based on deep residual network is proposed in this paper.Firstly,data enhancement,image resizing and TFRecord coding are used to preprocess the input of the network.Then a 34 layer deep residual network(ResNet-34) is designed to extract the characteristics of psoriasis.In addition,cross entropy is used as the loss function of resnet-34 to measure the accuracy of the model,and Adam algorithm is used as the optimizer to train ResNet-34.Finally an optimized ResNet-34 model is obtained for psoriasis diagnosis.The experimental results based on K-fold cross validation show that the proposed model is superior to the other diagnostic methods in terms of recall rate and ROC curve,and can provide technical support for psoriasis data analysis and disease prevention.
作者
李文胜
Li Wensheng(Hunan Vocational College of Electronic Science and technology,Changsha,Hunan 410220,China)
出处
《计算机时代》
2022年第7期1-6,共6页
Computer Era
基金
2021年湖南省教育厅科学研究项目“‘三教’改革背景下高职软件开发类混合式‘金课’的构建研究与实践”(21C1021)。
