[1]李碧草,王晶,郭旭伟,等.基于多尺度特征融合与反向注意力的COVID-19病灶分割[J].中国医学物理学杂志,2023,40(4):403-409.[doi:DOI:10.3969/j.issn.1005-202X.2023.04.002]
　LI Bicao,WANG Jing,GUO Xuwei,et al.COVID-19 lesion segmentation based on multi-scale feature fusion and reverse attention[J].Chinese Journal of Medical Physics,2023,40(4):403-409.[doi:DOI:10.3969/j.issn.1005-202X.2023.04.002]
点击复制

基于多尺度特征融合与反向注意力的COVID-19病灶分割()
分享到：

《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]

卷:
40卷
期数:
2023年第4期
页码:
403-409
栏目:
医学影像物理
出版日期:
2023-04-25

文章信息/Info

Title:
COVID-19 lesion segmentation based on multi-scale feature fusion and reverse attention
文章编号:
1005-202X（2023）04-0403-07
作者:
李碧草1; 2; 王晶1; 郭旭伟3; 黄杰1; 魏苗苗1; 李盼盼1
1.中原工学院电子信息学院， 河南 郑州 450007； 2.郑州大学信息工程学院， 河南 郑州 450001； 3.河南科技大学第一附属医院儿科， 河南 洛阳 471000
Author(s):
LI Bicao1; 2;  WANG Jing1;  GUO Xuwei3;  HUANG Jie1;  WEI Miaomiao1;  LI Panpan1
1. School of Electronic and Information, Zhongyuan University of Technology, Zhengzhou 450007, China 2. School of Information Engineering, Zhengzhou University, Zhengzhou 450001, China 3. Department of Pediatrics, the First Affiliated Hospital of Henan University of Science and Technology, Luoyang 471000, China
关键词:
COVID-19; 肺炎感染分割; 全局上下文聚合; 多尺度特征融合; 反向注意力
Keywords:
Keywords: COVID-19 lung lesion segmentation global context aggregation multi-scale feature fusion reverse attention
分类号:
R318；TP391
DOI:
DOI:10.3969/j.issn.1005-202X.2023.04.002
文献标志码:
A
摘要:
针对新型冠状病毒肺炎（COVID-19）分割问题中感染区域具有高变异性以及病灶与背景对比度低等问题，提出一种基于多尺度特征融合与反向注意力的COVID-19感染分割网络。首先，利用残差网络作为主干网络进行特征提取，并使用全局上下文聚合策略对不同层次特征进行融合得到粗略的分割结果；其次，在网络瓶颈处添加多尺度特征融合模块，利用空洞卷积与多核池化增强网络分割不同尺度病变的能力；最后，设计一种级联结构的反向注意力模块，利用互补区域的细节特征增强背景与目标的对比度。本文方法在COVID-19 CT分割测试集上的准确率、特异性、灵敏度分别达到0.714、0.700和0.958，误检和漏检区域明显减少，细小病灶的分割能力显著提升。
Abstract:
Abstract: A COVID-19 lesion segmentation network based on multi-scale feature fusion and reverse attention (MFFRA) is proposed to overcome the problems of high variability and low contrast between lesion and background in COVID-19 segmentation. The residual network is used as the backbone network to extract features, and the global context aggregation strategy is adopted to integrate different hierarchical features for obtaining rough segmentation results. In addition, the multi-scale feature fusion module is added at the bottleneck of the network to enable the ability to segment lesions of different scales using atrous convolutions and multi-kernel pooling. Finally, a novel cascaded reverse attention module is designed to improve the low contrast between normal tissue and lesions based on the detailed features of complementary regions. The proposed method has an accuracy, specificity and sensitivity of 0.714, 0.700, 0.958 on the COVID-19 CT test set, reduces the areas of misdetection and missed detection, and enhances the segmentation ability of fine lesions.

相似文献/References:

[1]刘发明,江桂华,杨宁,等.新型冠状病毒肺炎的影像组学研究[J].中国医学物理学杂志,2020,37(4):463.[doi:DOI:10.3969/j.issn.1005-202X.2020.04.012]
　LIU Faming,JIANG Guihua,YANG Ning,et al.Radiomics analysis on COVID-19[J].Chinese Journal of Medical Physics,2020,37(4):463.[doi:DOI:10.3969/j.issn.1005-202X.2020.04.012]
[2]邓靓娜,张斌,蒋健,等.窗宽和层厚在COVID-19磨玻璃征象中的诊断价值[J].中国医学物理学杂志,2020,37(5):619.[doi:10.3969/j.issn.1005-202X.2020.05.017]
　DENG Liangna,,et al.Diagnostic value of window width and slice thickness in ground glass opacity of COVID-19[J].Chinese Journal of Medical Physics,2020,37(4):619.[doi:10.3969/j.issn.1005-202X.2020.05.017]
[3]乔文俊,许乙凯,严承功,等.新型冠状病毒肺炎期间方舱CT感染防控经验[J].中国医学物理学杂志,2020,37(7):903.[doi:DOI:10.3969/j.issn.1005-202X.2020.07.020]
　QIAO Wenjun,XU Yikai,et al.Experience of COVID-19 prevention and control in shelter CT[J].Chinese Journal of Medical Physics,2020,37(4):903.[doi:DOI:10.3969/j.issn.1005-202X.2020.07.020]
[4]张子宇,赵可辉,牛慧芳,等.基于轻量级RG-DenseNet的COVID-19 CT图像分类[J].中国医学物理学杂志,2023,40(12):1494.[doi:DOI:10.3969/j.issn.1005-202X.2023.12.007]
　ZHANG Ziyu,ZHAO Kehui,NIU Huifang,et al.COVID-19 classification on CT image using lightweight RG DenseNet[J].Chinese Journal of Medical Physics,2023,40(4):1494.[doi:DOI:10.3969/j.issn.1005-202X.2023.12.007]
[5]齐瑜鹏,赵言龙,郑浩然.一种通过代谢差异分析抑制SARS-CoV-2复制的靶点预测方法[J].中国医学物理学杂志,2023,40(12):1577.[doi:DOI:10.3969/j.issn.1005-202X.2023.12.019]
　QI Yupeng,ZHAO Yanlong,ZHENG Haoran.Target prediction approach to inhibit SARS-CoV-2 replication based on metabolic difference analysis[J].Chinese Journal of Medical Physics,2023,40(4):1577.[doi:DOI:10.3969/j.issn.1005-202X.2023.12.019]

备注/Memo

备注/Memo:
【收稿日期】2022-12-26 【基金项目】国家自然科学基金（61901537）；河南省高校科技创新人才支持计划（23HASTIT030）；河南省留学人员科研择优项目；中国博士后科学基金（2020M672274）；中原工学院学科实力提升青年硕导培育计划（SD202207） 【作者简介】李碧草，博士，副教授，硕士生导师，主要研究方向：人工智能与医学图像处理，E-mail: lbc@zut.edu.cn
更新日期/Last Update: 2023-04-25