基于迁移学习的乳腺病理图片分类研究  

Research on Classification of Breast Pathology ImagesBased on Transfer Learning

作　　者：马春洁 MA Chunjie(School of Mathematics and Computer Application,Shangluo University,Shangluo 726000,China)

机构地区：[1]商洛学院,数学与计算机应用学院,陕西商洛726000

出　　处：《微型电脑应用》2024年第8期42-45,50,共5页Microcomputer Applications

基　　金：陕西省教育厅一般专项(22JK0364)。

摘　　要：为了辅助乳腺组织病理性切片的良恶性诊断,使用BreakHis公开数据集进行乳腺组织病理图片分类研究。考虑到数据的染色差异和数量不足,分别进行数据的染色归一化和数据增强预处理;采用迁移学习,将微调后的ResNet 50模型用于对预处理后的数据进行分类,得到最高分类准确率为99.60%的结果;将所得结果与现有研究结果进行比较,证明所提方法对处理乳腺组织图片分类问题具有高效性。To assist the benign and malignant diagnosis of pathological section of breast tissue,this study utilizes the BreakHis public dataset to carry out breast tissue pathology images classification.By considering the staining differences and insufficient quantity of data,the staining normalization and data augmentation preprocessing were respectively carried out.By transfer learning,the fine-tuned ResNet 50 model is used to classify the preprocessed data,which results in the highest classification accuracy is 99.60%.The obtained results are compared with existing research outcomes,and proved that the effectiveness of the proposed method in addressing breast tissue image classification.

关 键 词：乳腺病理图片 染色归一化 数据增强 图片分类 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]