基于双尺度并行注意力网络的肺实质分割  

Lung parenchyma segmentation based on double scale parallel attention network

作　　者：冯凯丽 任莉莉[2] 吴彦林 李艳 王洪瑞[1] 王光磊[1] FENG Kaili;REN Lili;WU Yanlin;LI Yan;WANG Hongrui;WANG Guanglei(School of Electronic Information Engineering,Hebei University,Baoding,Hebei 071002,P.R.China;Affiliated Hospital of Hebei University,Baoding,Hebei 071002,P.R.China)

机构地区：[1]河北大学电子信息工程学院,河北保定071002 [2]河北大学附属医院,河北保定071002

出　　处：《生物医学工程学杂志》2022年第4期721-729,共9页Journal of Biomedical Engineering

基　　金：国家自然科学基金项目(61473112);河北省自然科学基金重点项目(F2017201222)。

摘　　要：自动准确地对肺实质进行分割对于肺癌辅助诊断至关重要。近年来,深度学习领域的研究者们提出了许多基于U型网络(U-Net)改进的肺实质分割方法。但是现有的分割方法忽视了不同层级间特征图语义信息的融合互补,并且无法区分特征图中不同空间与通道的重要性。为解决该问题,本文提出双尺度并行注意力(DSPA)网络(DSPA-Net)架构,在“编码器—解码器”结构中引入了DSPA模块和空洞空间金字塔池化(ASPP)模块。其中,DSPA模块通过协同注意力(CA)得到特征图精确的空间和通道信息,并对不同层级特征图的语义信息进行聚合。ASPP模块利用不同空洞率的多个并行卷积核获取不同感受野下包含多尺度信息的特征图。两个模块分别解决了不同层级特征图与同一层级特征图中多尺度信息处理问题。本文在卡格尔(Kaggle)竞赛数据集上进行了实验验证,实验结果证明该网络架构与目前主流的分割网络相比具有明显的优势,戴斯相似性系数(DSC)和交并比(IoU)的值分别达到了0.972±0.002和0.945±0.004。基于以上研究,本文实现了肺实质自动准确的分割,为注意力机制和多尺度信息在肺实质分割领域的应用提供参考。Automatic and accurate segmentation of lung parenchyma is essential for assisted diagnosis of lung cancer.In recent years,researchers in the field of deep learning have proposed a number of improved lung parenchyma segmentation methods based on U-Net.However,the existing segmentation methods ignore the complementary fusion of semantic information in the feature map between different layers and fail to distinguish the importance of different spaces and channels in the feature map.To solve this problem,this paper proposes the double scale parallel attention(DSPA)network(DSPA-Net)architecture,and introduces the DSPA module and the atrous spatial pyramid pooling(ASPP)module in the“encoder-decoder”structure.Among them,the DSPA module aggregates the semantic information of feature maps of different levels while obtaining accurate space and channel information of feature map with the help of cooperative attention(CA).The ASPP module uses multiple parallel convolution kernels with different void rates to obtain feature maps containing multi-scale information under different receptive fields.The two modules address multi-scale information processing in feature maps of different levels and in feature maps of the same level,respectively.We conducted experimental verification on the Kaggle competition dataset.The experimental results prove that the network architecture has obvious advantages compared with the current mainstream segmentation network.The values of dice similarity coefficient(DSC)and intersection on union(IoU)reached 0.972±0.002 and 0.945±0.004,respectively.This paper achieves automatic and accurate segmentation of lung parenchyma and provides a reference for the application of attentional mechanisms and multi-scale information in the field of lung parenchyma segmentation.

关 键 词：肺实质分割 深度学习 协同注意力 多尺度信息 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] R734.2[自动化与计算机技术—计算机科学与技术]