欢迎访问《图学学报》 分享到:

图学学报 ›› 2023, Vol. 44 ›› Issue (1): 50-58.DOI: 10.11996/JG.j.2095-302X.2023010050

• 图像处理与计算机视觉 • 上一篇    下一篇

融合注意力机制的肠道息肉分割多尺度卷积神经网络

单芳湄1(), 王梦文1, 李敏1,2()   

  1. 1.南京理工大学计算机科学与工程学院,江苏 南京 210094
    2.山东省数字医学与计算机辅助手术重点实验室,山东 青岛 266003
  • 收稿日期:2022-03-24 修回日期:2022-08-05 出版日期:2023-10-31 发布日期:2023-02-16
  • 通讯作者: 李敏
  • 作者简介:单芳湄(1997-),女,硕士研究生。主要研究方向为图像分割。E-mail:1094264762@qq.com
  • 基金资助:
    国家自然科学基金项目(61501241);江苏省自然科学基金项目(BK20150792);山东省数字医学与计算机辅助手术重点实验室开放基金项目(SDKL-DMCAS-2018-04);江苏省交通运输科技项目(2021Y)

Multi-scale convolutional neural network incorporating attention mechanism for intestinal polyp segmentation

SHAN Fang-mei1(), WANG Meng-wen1, LI Min1,2()   

  1. 1. School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing Jiangsu 210094, China
    2. Shandong Key Laboratory of Digital Medicine and Computer Assisted Surgery, Qingdao Shandong 266003, China
  • Received:2022-03-24 Revised:2022-08-05 Online:2023-10-31 Published:2023-02-16
  • Contact: LI Min
  • About author:SHAN Fang-mei (1997-), master student. Her main research interest covers image segmentation. E-mail:1094264762@qq.com
  • Supported by:
    National Natural Science Foundation of China(61501241);National Natural Science Foundation of Jiangsu Province(BK20150792);Foundation of Shandong Provincial Key Laboratory of Digital Medicine and Computer Assisted Surgery(SDKL-DMCAS-2018-04);Transport Science and Technology Project of Jiangsu Province(2021Y)

摘要:

肠道息肉分割能够提供息肉在结肠中的位置和形态信息,方便医生依据其结构变化程度来推断癌变可能性,有利于结肠癌的早期诊断和治疗。针对许多现有的卷积神经网络所提取的多尺度特征有限,且常引入冗余和干扰特征,难以应对复杂多变的肠道息肉分割问题,提出了一种融合注意力机制的肠道息肉分割多尺度卷积神经网络(CNN)。首先,设计不同比例金字塔池化策略提取丰富的多尺度上下文信息;然后,通过在网络中融入通道注意力机制,模型能够根据目标自适应地选择合适的局部上下文信息和全局上下文信息进行特征集成;最后,联合金字塔池化策略和通道注意力机制构建多尺度有效语义融合解码网络,增强模型对形状、大小复杂多变的肠道息肉分割的鲁棒性。实验结果表明,本文模型分割的Dice系数、IoU和灵敏度在CVC-ClinicDB数据集上分别为90.6%,84.4%和91.1%,在ETIS-Larib数据集上分别为80.6%,72.6%和79.0%,其能够从肠镜图像中准确、有效地分割出肠道息肉。

关键词: 息肉分割, 肠镜图像, 卷积神经网络, 多尺度语义信息

Abstract:

Intestinal polyp segmentation provides the location and morphology of polyps in colon, allowing doctors to infer the possibility of canceration according to the degree of structural deformation, which facilitates the early diagnosis and treatment of colon cancer. In view of the limited multi-scale features extracted by many existing convolutional neural networks (CNN), and the frequently caused redundant and interfering features, it is difficult to extract the complex and variable targets. To address this challenge, a multi-scale convolutional neural network incorporating attention mechanism was proposed for intestinal polyp segmentation. Specifically, the pyramid strategy based on different scales of pooling was designed to capture the rich multi-scale context information. Then a channel attention mechanism was incorporated into the network so that the model could adaptively select appropriate local and global contextual information for feature integration based on the region of interest. Following that, by combining the pyramid pooling strategy and the channel attention mechanism, a multi-scale effective semantic fusion decoder network was constructed to improve the model robustness for segmentation of intestinal polyps with complex and variable shapes and sizes. The experimental results show that the Dice coefficient, IoU, and sensitivity produced by the proposed model reach 90.6%, 84.4%, and 91.1% on the CVC-ClinicDB dataset, and 80.6%, 72.6%, and 79.0% on the ETIS-Larib dataset, indicating that the proposed model could accurately and effectively segments polyps in colonoscopy images.

Key words: polyp segmentation, colonoscopy images, convolutional neural network, multiscale semantic information

中图分类号: