计算机应用 ›› 2017, Vol. 37 ›› Issue (5): 1434-1438.DOI: 10.11772/j.issn.1001-9081.2017.05.1434

• 计算机视觉与虚拟现实 • 上一篇    下一篇

基于深度学习的超声心动图切面识别方法

陶攀1,2, 付忠良1,2, 朱锴1,2, 王莉莉1,2   

  1. 1. 中国科学院 成都计算机应用研究所, 成都 610041;
    2. 中国科学院大学, 北京 100049
  • 收稿日期:2016-09-27 修回日期:2016-12-30 出版日期:2017-05-10 发布日期:2017-05-16
  • 通讯作者: 陶攀
  • 作者简介:陶攀(1988-),男,河南安阳人,博士研究生,主要研究方向:机器学习、医学图像处理;付忠良(1967-),男,重庆人,研究员,博士生导师,主要研究方向:机器学习、数据挖掘;朱锴(1991-),男,贵州安顺人,博士研究生,主要研究方向:机器学习;王莉莉(1987-),女,河南周口人,博士研究生,主要研究方向:机器学习。
  • 基金资助:
    中国科学院西部之光人才培养计划项目。

Echocardiogram view recognition using deep convolutional neural network

TAO Pan1,2, FU Zhongliang1,2, ZHU Kai1,2, WANG Lili1,2   

  1. 1. Chengdu Institute of Computer Application, Chinese Academy of Sciences, Chengdu Sichuan 610041, China;
    2. University of Chinese Academy of Sciences, Beijing 100049, China
  • Received:2016-09-27 Revised:2016-12-30 Online:2017-05-10 Published:2017-05-16
  • Supported by:
    This work is supported by the the Project of West Light Foundation of the Chinese Academy of Sciences (R&D and Application of Cardiac Function Evaluation System based on Medical Image Modeling).

摘要: 提出了一种基于深度卷积神经网络自动识别超声心动图标准切面的方法,并可视化分析了深度模型的有效性。针对网络全连接层占有模型大部分参数的缺点,引入空间金字塔均值池层化替代全连接层,获得更多空间结构信息,并大大减少模型参数、降低过拟合风险,通过类别显著性区域将类似注意力机制引入模型可视化过程。通过超声心动图标准切面的识别问题案例,对深度卷积神经网络模型的鲁棒性和有效性进行解释。在超声心动图上的可视化分析实验表明,改进深度模型作出的识别决策依据,同医师辨别分类超声心动图标准切面的依据一致,表明所提方法的有效性和实用性。

关键词: 深度学习, 标准切面分类, 超声心动图, 可视化, 卷积神经网络

Abstract: A deep model for automatic recognition of echocardiographic standard views based on deep convolutional neural network was proposed, and the effectiveness of the deep model was analyzed by visualize class activation maps. In order to overcome the shortcomings of the fully connected layer occupying most of the parameters of the model, the spatial pyramid mean pool was used to replace the fully connected layer, and more spatial structure information was obtained. The model parameters and the over-fitting risk were reduced.The attention mechanism was introduced into the model visualization process by the class significance region. The robustness and effectiveness of the deep convolution neural network model were explained by the case recognizing echocardiographic standard views. Visualization analysis on echocardiography show that the decision basis made by the improved depth model is consistent with the standard view classification by the sonographer which indicates the validity and practicability of the proposed method.

Key words: deep learning, standard view classification, echocardiogram, visualization, Convolutional Neural Network (CNN)

中图分类号: