计算机科学与探索 ›› 2022, Vol. 16 ›› Issue (8): 1885-1897.DOI: 10.3778/j.issn.1673-9418.2011091

• 图形图像 • 上一篇    下一篇

权重初始化-滑动窗口CNN的医学图像分类

安凤平1,+(), 李晓薇1, 曹翔1   

  1. 1. 淮阴师范学院 物理与电子电气工程学院,江苏 淮安 223300
    2. 北京理工大学 信息与电子学院,北京 100081
  • 收稿日期:2020-11-27 修回日期:2021-01-29 出版日期:2022-08-01 发布日期:2021-03-03
  • 通讯作者: +E-mail: anfengping@163.com
  • 作者简介:安凤平(1985—),男,安徽芜湖人,博士,副教授,硕士生导师,主要研究方向为图像处理、深度学习。
    李晓薇(1962—),女,江苏扬州人,硕士,教授,硕士生导师,主要研究方向为超导隧道结、混杂系统。
    曹翔(1981—),男,四川荣县人,博士,副教授,硕士生导师,主要研究方向为多水下机器人协作控制、水下机器人任务规划。
  • 基金资助:
    国家自然科学基金(61701188);江苏省自然科学基金(BK20201479);中国博士后科学基金(2019M650512)

Medical Image Classification Algorithm Based on Weight Initialization-Sliding Window CNN

AN Fengping1,+(), LI Xiaowei1, CAO Xiang1   

  1. 1. School of Physics and Electronic Electrical Engineering, Huaiyin Normal University, Huai’an, Jiangsu 223300, China
    2. School of Information and Electronics, Beijing Institute of Technology, Beijing 100081, China
  • Received:2020-11-27 Revised:2021-01-29 Online:2022-08-01 Published:2021-03-03
  • About author:AN Fengping, born in 1985, Ph.D., associate professor, M.S. supervisor. His research interests include image processing and deep learning.
    LI Xiaowei, born in 1962, M.S., professor, M.S. supervisor. Her research interests include superconducting tunnel junction and hybrid system.
    CAO Xiang, born in 1981, Ph.D., associate professor, M.S. supervisor. His research interests include cooperative control of multiple underwater robots and task planning of underwater robots.
  • Supported by:
    This work was supported by the National Natural Science Foundation of China(61701188);the Natural Science Foundation of Jiangsu Province(BK20201479);the Postdoctoral Science-Foundation of China(2019M650512)

摘要:

深度学习在医学图像分类应用过程中存在以下问题:一是无法针对医学图像性质构建深度学习模型层级;二是深度学习模型网络初始化权重未能得到较好优化。为此,首先从网络优化角度出发,通过优化方法提高网络的非线性建模能力,提出了一种新的网络权重初始化方法,缓解了现有深度学习的初始化理论受限于非线性单元类型的问题,增加了神经网络处理不同视觉任务的潜力。同时,为了充分利用医学图像的特性,通过对多列卷积神经网络框架进行深入研究,发现通过改变卷积神经网络不同层次的特征数目和卷积核大小,可以构建不同的卷积神经网络模型,以更好地适应待处理医学图像的医学特性,并训练得到的异构多列卷积神经网络。最后,利用提出的一种自适应的滑动窗口融合机制,共同完成医学图像的分类任务。基于上述思想,提出了一种基于权重初始化-多层卷积神经网络滑动窗口融合的医学分类算法。利用提出方法对乳腺肿块分类、脑肿瘤组织分类实验和医学图像数据库分类分别进行实验,实验结果表明,所提方法不仅平均准确率较传统机器学习、其他深度学习方法有明显提高,而且具有较好的稳定性和鲁棒性。

关键词: 深度学习, 权重初始化, 滑动窗口, 卷积神经网络(CNN), 医学图像分类

Abstract:

Deep learning has the following problems in medical image classification application: first, it is impossible to construct a deep learning model hierarchy for medical image properties; second, the network initialization weight of the deep learning model has not been optimized. To this end, this paper starts from the perspective of network optimization, and then improves the nonlinear modeling ability of the network through optimization methods. Then, this paper proposes a new network weight initialization method, which alleviates the problem that the initialization theory of existing deep learning is limited by the nonlinear unit type, and increases the potential of neural network to deal with different visual tasks. At the same time, in order to make full use of the characteristics of medical images, this paper deeply studies the multi-column convolutional neural network framework and finds that through changing the number of features and the convolution kernel size of different levels of convolutional neural networks, it can construct different convolutional neural network models to better adapt to the medical characteristics of the medical images to be processed and train the obtained heterogeneous multi-column convolutional neural networks. Finally, the classification task of medical images is completed by the method proposed in this paper. Based on the above ideas, this paper proposes a medical classification algorithm based on weight initialization-sliding window fusion of multi-layer convolutional neural networks. The methods of this paper are used to classify breast mass classification, brain tumor tissue classification experiment and medical image database classification. The experimental results show that the proposed method not only has higher average accuracy than traditional machine learning and other deep learning methods, but also has better stability and robustness.

Key words: deep learning, weight initialization, sliding window, convolutional neural network (CNN), medical image classification

中图分类号: