基于卷积神经网络的胃癌癌前病变图像分类方法  

Image classification of gastric precancerous lesions based on convolutional neural network

作　　者：张育 赵轶峰[1] 苏卓彬[1] 杨永江[1] ZHANG Yu;ZHAO Yifeng;SU Zhuobin;YANG Yongjiang(Department of Gastrointestinal Tumor Surgery,the First Affiliated Hospital of Hebei North University,Zhangjiakou 075000,China)

机构地区：[1]河北北方学院附属第一医院胃肠肿瘤外科,河北张家口075000

出　　处：《中国医学物理学杂志》2022年第2期209-214,共6页Chinese Journal of Medical Physics

基　　金：河北省卫生厅科研基金(20200502)。

摘　　要：通过建立一个以系统和智能方式对胃癌癌前病变进行分类的模型,帮助医生找到敏感点和癌前息肉。在本文方法中,通过设计一种改进的ALexNet架构并使用数据增强、高斯噪声、L2权值衰减和ReLU等技术训练卷积神经网络模型,最后通过利用精度、损失值和混淆矩阵等性能指标对该模型的性能进行评估。在3677张糜烂、息肉和溃疡等胃病图像上对所提出的模型进行测试,结果表明该模型的分类准确率达到89%。By building a model that can classify gastric precancerous lesions in a systematic and intelligent way,doctors can find sensitive points and precancerous polyps.An improved AlexNet architecture and techniques such as data enhancement,Gaussian noise,L2 weight decay and ReLU are used for training the convolutional neural network model;and the performance of the proposed model is evaluated by analyzing its precision,loss value and confusion matrix.The proposed model is tested on 3677 images of gastric diseases such as erosion,polyps and ulcers,and the results show that the classification accuracy of the proposed model reaches 89%.

关 键 词：胃癌 糜烂 息肉 溃疡 癌前病变 卷积神经网络 图像分类 

分 类 号：R318[医药卫生—生物医学工程] TP391[医药卫生—基础医学]