Journal of Computer Applications ›› 2021, Vol. 41 ›› Issue (4): 1221-1226.DOI: 10.11772/j.issn.1001-9081.2020071034

Special Issue: 前沿与综合应用

• Frontier and comprehensive applications • Previous Articles    

Cerebral infarction image recognition based on semi-supervised method

OU Lili1, SHAO Fengjing1, SUN Rencheng1,2, SUI Yi1   

  1. 1. College of Computer Science and Technology, Qingdao University, Qingdao Shandong 266071, China;
    2. The Affiliated Hospital of Qingdao University, Qingdao Shandong 266071, China
  • Received:2020-07-17 Revised:2020-10-09 Online:2021-04-10 Published:2020-11-12
  • Supported by:
    This work is partially supported by the Youth Program of National Natural Science Foundation of China (41706198).

基于半监督方法的脑梗死图像识别

欧莉莉1, 邵峰晶1, 孙仁诚1,2, 隋毅1   

  1. 1. 青岛大学 计算机科学技术学院, 山东 青岛 266071;
    2. 青岛大学附属医院, 山东 青岛 266071
  • 通讯作者: 邵峰晶
  • 作者简介:欧莉莉(1993—),女,河南商丘人,硕士研究生,主要研究方向:大数据分析;邵峰晶(1955—),女,山东青州人,教授,博士生导师,博士,主要研究方向:数据挖掘、复杂网络;孙仁诚(1977—),男,山东即墨人,副教授,博士,主要研究方向:数据挖掘、复杂网络;隋毅(1984—),女,山东青岛人,副教授,博士,主要研究方向:大数据建模与分析、复杂网络、机器学习。
  • 基金资助:
    国家自然科学基金青年科学基金资助项目(41706198)。

Abstract: In the field of image recognition, images with insufficient label data cannot be well recognized by the supervised method model. In order to solve this problem, a semi-supervised method model based on Generative Adversarial Network(GAN) was proposed. That is, by combining the advantages of semi-supervised GANs and deep convolutional GANs, and replacing the sigmoid activation function with softmax in the output layer, the Semi-Supervised Deep Convolutional GAN(SS-DCGAN) model was established. Firstly, the generated samples were defined as pseudo-samples and used to guide the training process. Secondly, the semi-supervised training method was adopted to update the parameters of the model. Finally, the recognition of abnormal(cerebral infarction) images was realized. Experimental results show that the SS-DCGAN model can recognize abnormal images well with little label data, which achieves 95.05% recognition rates. Compared with Residual Network 32(ResNet32) and Ladder networks, the SS-DCGAN model has significant advantages.

Key words: Generative Adversarial Network (GAN), semi-supervised, cerebral infarction, deep convolutional networks, image recognition, feature matching

摘要: 在图像识别领域,针对有监督方法的模型在标签数据不足时图像的识别效果不佳问题,提出一种基于生成对抗网络(GAN)的半监督方法模型,即结合了半监督生成对抗网络(SSGAN)和深度卷积生成对抗网络(DCGAN)的优点,并在输出层用softmax代替了sigmoid激活函数,从而建立半监督深度卷积生成对抗网络(SS-DCGAN)模型。首先,将生成样本定义为伪样本类别并用于引导训练;其次,采用半监督的训练方式对模型的参数进行更新;最后,实现对异常(脑梗死)图像的识别。实验结果表明,SS-DCGAN模型在标签数据较少时能够很好地识别异常图像,达到95.05%的识别率,与ResNet32、半监督梯度网络(Ladder Network)分类方法相比具有显著的优越性。

关键词: 生成对抗网络, 半监督, 脑梗, 深度卷积网络, 图像识别, 特征匹配

CLC Number: