东北大学学报:自然科学版 ›› 2020, Vol. 41 ›› Issue (7): 1054-1061.DOI: 10.12068/j.issn.1005-3026.2020.07.022

• 生物工程 • 上一篇    下一篇

基于生成对抗网络的低分化宫颈癌病理图像分类

李晨1, 张家伟1, 张昊1, 汪茜2,3   

  1. (1.东北大学 医学与生物信息工程学院, 辽宁 沈阳110169; 2.辽宁省肿瘤医院, 辽宁 沈阳110042;3.中国医科大学附属肿瘤医院, 辽宁 沈阳110042)
  • 收稿日期:2019-07-30 修回日期:2019-07-30 出版日期:2020-07-15 发布日期:2020-07-15
  • 通讯作者: 李晨
  • 作者简介:李晨(1985-),男,辽宁沈阳人,东北大学副教授.
  • 基金资助:
    国家自然科学基金资助项目(61806047,81902676); 中央高校基本科研业务费专项资金资助项目(N2019003).

Generative Adversarial Networks Based Pathological Images Classification of Poorly Differentiated Cervical Cancer

LI Chen1, ZHANG Jia-wei1, ZHANG Hao1, WANG Qian2,3   

  1. 1. College of Medicine and Biological Information Engineering, Northeastern University, Shenyang 110169, China; 2. Liaoning Cancer Hospital & Institute, Shenyang 110042, China; 3.Cancer Hospital of China Medical University, Shenyang 110042, China.
  • Received:2019-07-30 Revised:2019-07-30 Online:2020-07-15 Published:2020-07-15
  • Contact: WANG Qian
  • About author:-
  • Supported by:
    -

摘要: 使用生成对抗网络(GAN)扩充宫颈癌病理图像的数据集以提高计算机辅助诊断的准确率.首先,使用GAN进行细胞质部分图像生成;其次,使用两次k-means聚类对生成图像进行筛选;最后,使用Inception-V3模型对数据集进行分类训练.结果表明,在测试集相同的情况下,该方法可以将总体分类准确率提升约2.5%,尤其对低分化宫颈癌病理图像有显著效果.通过GAN解决了组织病理学图像无方向性、内容复杂、前景目标规则性差等问题,证明了该方法的有效性及发展潜力.

关键词: 宫颈癌辅助诊断, 组织病理学图像分类, 生成对抗网络, 特征提取, k-means聚类

Abstract: The accuracy of computer-assisted diagnosis can be improved by using generative adversarial networks(GAN) to extend the data set of cervical cancer patholigical images. First, the cytoplasmic part of the histopathological images was generated by GAN; then, k-means clustering was used twice to select images generated by GAN; finally, Inception-V3 model was used to train a classifier. The results showed that the accuracy is improved by an average of 2.5% under the same test data set. Especially, it has significant effect for poorly differentiated cervical cancer pathological images. The non-directionality, complexity of content and poor regularity of foreground target for histopathological images are solved by GAN, which proves the effectiveness and the potential of this method.

Key words: auxiliary diagnosis for cervical cancer, histopathological image classification, generative adversarial networks(GAN), feature extracting, k-means clustering

中图分类号: