[1]岳丹鸣,汪雪林,邓神谧,等.基于可形变编码器-解码器网络的视网膜血管精准分割[J].中国医学物理学杂志,2026,43(7):866-878.[doi:DOI:10.3969/j.issn.1005-202X.2026.07.006]
 YUE Danming,WANG Xuelin,DENG Shenmi,et al.Deformable encoder-decoder network for accurate retinal vessel segmentation[J].Chinese Journal of Medical Physics,2026,43(7):866-878.[doi:DOI:10.3969/j.issn.1005-202X.2026.07.006]
点击复制

基于可形变编码器-解码器网络的视网膜血管精准分割()

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

卷:
43卷
期数:
2026年第7期
页码:
866-878
栏目:
医学影像物理
出版日期:
2026-07-22

文章信息/Info

Title:
Deformable encoder-decoder network for accurate retinal vessel segmentation
文章编号:
1005-202X(2026)07-0866-13
作者:
岳丹鸣1汪雪林2邓神谧1阮雍硕1江静1
1.北京联合大学智慧城市学院, 北京 100101;2.青岛城市学院机电工程学院, 山东 青岛 266106
Author(s):
YUE Danming1 WANG Xuelin2 DENG Shenmi1 RUAN Yongshuo1 JIANG Jing1
1. Smart City College, Beijing Union University, Beijing 100101, China 2. School of Mechanical and Electrical Engineering, Qingdao City University, Qingdao 266106, China
关键词:
医学图像视网膜血管分割残差可形变卷积分组聚合桥接轻量化网络
Keywords:
medical image retinal vessel segmentation residual deformable convolution group aggregation bridge lightweight network
分类号:
R318;TP391
DOI:
DOI:10.3969/j.issn.1005-202X.2026.07.006
文献标志码:
A
摘要:
针对现有方法在视网膜血管分割中,捕获血管不规则形态及多尺度特征融合方面的不足,提出一种轻量化可形变的编码器-解码器视网膜血管分割网络ResDC-Net,该网络由编码器、跳跃连接模块和解码器组成。编码器和解码器部分主要由残差可形变卷积模块构成,该模块利用可形变卷积动态适配血管的形态,并通过残差连接增强特征流,有效解决传统卷积对弯曲血管、分支结构的表征不足问题,提升血管形变的建模能力。运用分组聚合桥接模块作为跳跃连接模块,该模块融合编码器输出的不同深度的信息和解码器在每个阶段生成的掩码信息,强化语义信息与空间细节的互补性,有助于恢复血管边缘细节。在STARE、CHASE_DB1、HRF(分辨率为876×584、1 752×1 168)公开数据集上进行实验,其中特异性分别达到98.66%、98.67%、98.43%和98.42%,准确率分别达到97.47%、97.63%、97.18%和97.17%。实验表明本文网络在视网膜血管分割任务中的有效性、轻量性与优越性。
Abstract:
To overcome the drawbacks of existing approaches in capturing irregular vascular morphologies and integrating multi-scale features, this study proposes a lightweight deformable encoder-decoder network named ResDC-Net for retinal vessel segmentation. The network consists of 3 components: an encoder, skip connection modules, and a decoder. Both the encoder and decoder are mainly composed of residual deformable convolution modules. These modules utilize deformable convolutions to dynamically adapt to vascular morphology and enhance feature propagation through residual connections, which effectively compensates for the limited representation capacity of traditional convolutions for curved vessels and branches, thereby improving the modeling for vascular deformations. A group aggregation bridge module serves as the skip connection module. This module integrates multi-depth information output by the encoder and the mask information generated by the decoder at each stage, strengthening the complementarity between semantic information and spatial details, which is conducive to restoring vessel edge details. Experiments on the public STARE, CHASE_DB1, and HRF (resolution: 876×584, 1 752×1 168) datasets show that the proposed method achieves specificity of 98.66%, 98.67%, 98.43% and 98.42%, and accuracy of 97.47%, 97.63%, 97.18% and 97.17%, respectively, demonstrating its effectiveness, lightweight property, and superiority for retinal vessel segmentation.

相似文献/References:

[1]张 硕,袁克虹.一种基于SSED的无规则噪声组织的去除方法[J].中国医学物理学杂志,2015,32(01):48.[doi:10.3969/j.issn.1005-202X.2015.01.012]
[2]赵 雷,唐 峰,鲁 雯,等.应用IQF值评价不同显示器的性能[J].中国医学物理学杂志,2014,31(03):4897.[doi:10.3969/j.issn.1005-202X.2014.03.012]
[3]杨 娟,陈兆学,姜 雪,等.基于Mean Shift的三维医学图像交互式分割方法[J].中国医学物理学杂志,2014,31(04):5026.[doi:10.3969/j.issn.1005-202X.2014.04.012]
[4]王双玲,龙成章,刘辉,等.基于双向异步投票策略医学图像特征点匹配[J].中国医学物理学杂志,2016,33(4):423.[doi:10.3969/j.issn.1005-202X.2016.04.020]
 [J].Chinese Journal of Medical Physics,2016,33(7):423.[doi:10.3969/j.issn.1005-202X.2016.04.020]
[5]徐效文,杨志刚.光学相干层析医学图像分割研究现状[J].中国医学物理学杂志,2016,33(7):697.[doi:10.3969/j.issn.1005-202X.2016.07.011]
 [J].Chinese Journal of Medical Physics,2016,33(7):697.[doi:10.3969/j.issn.1005-202X.2016.07.011]
[6]张光玉,鲁雯,徐龙春,等.DICOM格式医学图像显示方法研究[J].中国医学物理学杂志,2017,34(1):32.[doi:10.3969/j.issn.1005-202X.2017.01.007]
 [J].Chinese Journal of Medical Physics,2017,34(7):32.[doi:10.3969/j.issn.1005-202X.2017.01.007]
[7]王小东,赵仁宏,季超,等.基于Sobel算子的医学图像边缘提取算法的DSP编程设计与实现[J].中国医学物理学杂志,2017,34(7):690.[doi:10.3969/j.issn.1005-202X.2017.07.008]
 [J].Chinese Journal of Medical Physics,2017,34(7):690.[doi:10.3969/j.issn.1005-202X.2017.07.008]
[8]赵帅,吕晓琪.基于局部互信息和自动更新的医学图像形变不均匀配准806[J].中国医学物理学杂志,2017,34(8):806.[doi:DOI:10.3969/j.issn.1005-202X.2017.08.010]
[9]肖文君,刘伍丰,张建川. 基于西门子CT-SIM的医学图像研究[J].中国医学物理学杂志,2018,35(3):285.[doi:DOI:10.3969/j.issn.1005-202X.2018.03.008]
 XIAO Wenjun,LIU Wufeng,ZHANG Jianchuan. Medical image study based on Siemens CT-SIM[J].Chinese Journal of Medical Physics,2018,35(7):285.[doi:DOI:10.3969/j.issn.1005-202X.2018.03.008]
[10]孙畅,张超. 基于局部与非局部医学图像降噪技术的研究进展[J].中国医学物理学杂志,2018,35(5):565.[doi:DOI:10.3969/j.issn.1005-202X.2018.05.013]
 SUN Chang,ZHANG Chao. Progress in medical image denoising technology based on local and nonlocal filtering[J].Chinese Journal of Medical Physics,2018,35(7):565.[doi:DOI:10.3969/j.issn.1005-202X.2018.05.013]

备注/Memo

备注/Memo:
【收稿日期】2026-02-10 【基金项目】国家自然科学基金(61871055) 【作者简介】岳丹鸣,硕士研究生,研究方向:计算机视觉、医学图像处理,E-mail: 2856042349@qq.com 【通信作者】江静,博士,副教授,研究方向:数字图像处理、大数据分析,E-mail: xxtjiangjing@buu.edu.cn
更新日期/Last Update: 2026-07-22