高级搜索

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于三维模糊连接脉冲耦合神经网络的肾脏CT图像自动分割算法

白培瑞 李峥 刘庆一 王梦 毕丽君 任延德 王成健

白培瑞, 李峥, 刘庆一, 王梦, 毕丽君, 任延德, 王成健. 基于三维模糊连接脉冲耦合神经网络的肾脏CT图像自动分割算法[J]. 电子与信息学报, 2023, 45(6): 2264-2272. doi: 10.11999/JEIT221252
引用本文: 白培瑞, 李峥, 刘庆一, 王梦, 毕丽君, 任延德, 王成健. 基于三维模糊连接脉冲耦合神经网络的肾脏CT图像自动分割算法[J]. 电子与信息学报, 2023, 45(6): 2264-2272. doi: 10.11999/JEIT221252
BAI Peirui, LI Zheng, LIU Qingyi, WANG Meng, BI Lijun, REN Yande, WANG Chengjian. Automatic Kidney CT Images Segmentation Algorithm Based on 3D Fuzzy Connectedness and Pulse Coupled Neural Network[J]. Journal of Electronics & Information Technology, 2023, 45(6): 2264-2272. doi: 10.11999/JEIT221252
Citation: BAI Peirui, LI Zheng, LIU Qingyi, WANG Meng, BI Lijun, REN Yande, WANG Chengjian. Automatic Kidney CT Images Segmentation Algorithm Based on 3D Fuzzy Connectedness and Pulse Coupled Neural Network[J]. Journal of Electronics & Information Technology, 2023, 45(6): 2264-2272. doi: 10.11999/JEIT221252

基于三维模糊连接脉冲耦合神经网络的肾脏CT图像自动分割算法

doi: 10.11999/JEIT221252
基金项目: 国家自然科学基金(61471225)
详细信息
    作者简介:

    白培瑞:男,博士,教授,研究方向为医学成像技术、医学图像分析、视觉跟踪、模式识别和计算机视觉

    李峥:女,硕士生,研究方向为医学图像处理

    刘庆一:男,博士,讲师,研究方向为医学图像处理

    王梦:女,硕士生,研究方向为医学图像处理

    毕丽君:女,博士,讲师,研究方向为现代成像技术、智能传感器

    任延德:男,博士,主任医师,研究方向为神经系统疾病的影像诊断、深度学习在神经医学影像中的应用

    王成健:男,硕士,住院医师,研究方向为神经成像诊断、医学图像处理

    通讯作者:

    刘庆一 lqy_raphael@163.com

  • 中图分类号: TN911.73;TP391.41

Automatic Kidney CT Images Segmentation Algorithm Based on 3D Fuzzy Connectedness and Pulse Coupled Neural Network

Funds: The National Natural Science Foundation of China (61471225)
  • 摘要: 3维肾脏CT图像的自动准确分割对减轻医师阅片工作量和提高计算机辅助诊断效率具有重要意义。但是,由于肾脏器官的结构复杂性以及邻近部位的灰度相似性,3维肾脏的准确分割仍具有挑战性。该文基于简化脉冲耦合神经网络(SPCNN)结构简单、参数量少的特点,结合模糊连接度(FC)算法,提出一种3维肾脏CT图像的自动分割算法。主要贡献为:将SPCNN的2维模型扩展为3维模型,可以充分利用3维CT图像的层间信息;提出了一种基于感兴趣区域质心的3维种子点自动生成策略,可以有效提高算法的自动分割效率;实现了3维FC响应图与3维SPCNN的有效耦合。所提算法在自制数据集和公开数据集上进行了验证实验,结果表明该算法的性能优于现有的主流算法,其Dice系数、准确率、敏感度、体积误差、平均对称表面距离的平均值分别可以达到0.9095, 0.9969, 0.8517, 0.1749和0.8536。
  • 图  1  3D SPCNN神经元模型

    图  2  融合机制

    图  3  左右肾脏的种子点示意

    图  4  本文算法流程图

    图  5  肾脏分割结果比较

    图  6  肾脏的3维结果显示

    图  7  参数设置不同结果对比

    图  8  与3D U-Net分割结果对比

    表  1  3D FC-SPCNN算法与不同算法在AHQU_K数据集上的分割结果比较

    算法DiceAccSenVE↓ASSD↓
    3D FC-SPCNN0.9175±0.01450.9969±0.00050.8533±0.02990.1749±0.05050.7903±0.1480
    3D FC0.9003±0.03960.9964±0.00120.8268±0.06580.2213±0.07230.9552±0.4141
    3D 区域生长0.8916±0.05890.9954±0.00580.8252±0.05010.3151±0.14421.2365±0.5071
    3D Snakes0.8058±0.06870.9928±0.00290.7369±0.10280.2374±0.10361.6287±0.9566
    3D SPCNN0.1481±0.03870.7433±0.06340.7376±0.065613.6017±3.333220.1676±3.8033
    下载: 导出CSV

    表  2  3D FC-SPCNN算法与不同算法在3Dircadb数据集上的分割结果比较

    算法DiceAccSenVE↓ASSD↓
    3D FC-SPCNN0.9119±0.02270.9972±0.00070.8519±0.02270.1865±0.02930.8107±0.1505
    3D FC0.8738±0.07690.9964±0.00190.7926±0.10110.2492±0.09621.0926±0.5016
    3D 区域生长0.8620±0.08160.9935±0.00530.8192±0.06350.2874±0.13311.2059±0.6009
    3D Snakes0.8063±0.02230.9863±0.00380.7586±0.03820.2206±0.02481.4037±0.1379
    3D SPCNN0.2208±0.05610.9182±0.24250.7027±0.05612.2117±1.427919.4560±5.1872
    下载: 导出CSV

    表  3  3D FC-SPCNN算法与不同算法在Kits19数据集上的分割结果比较

    算法DiceAccSenVE↓ASSD↓
    3D FC-SPCNN0.8991±0.01030.9966±0.00140.8499±0.03050.1633±0.07400.9598±0.2463
    3D FC0.8383±0.03700.9907±0.00880.8136±0.05900.3392±0.20891.5468±0.6721
    3D 区域生长0.8621±0.05160.9949±0.00390.7805±0.10100.3197±0.13571.6331±0.8665
    3D Snakes0.8329±0.05260.9908±0.00410.7495±0.14050.2083±0.09161.3861±0.3569
    3D SPCNN0.0487±0.02060.5312±0.08310.7392±0.057330.5897±7.706223.914±4.0112
    下载: 导出CSV

    表  4  相同参数、种子点个数不同的分割结果的评价指标对比

    DiceAccSenVE↓ASSD↓时间(s)
    Z/20.91970.99670.85510.18640.81503172
    Z/2和Z/2–50.92080.99680.85720.18170.79193193
    Z/2和Z/2–5和Z/2+50.92130.99680.85810.18040.78633200
    下载: 导出CSV
  • [1] VAZIRI N D. Silva's diagnostic renal pathology[J]. Kidney International, 2010, 77(11): 939–940. doi: 10.1038/ki.2009.392
    [2] DOI K. Computer-aided diagnosis in medical imaging: Historical review, current status and future potential[J]. Computerized Medical Imaging and Graphics, 2007, 31(4/5): 198–211. doi: 10.1016/j.compmedimag.2007.02.002
    [3] TORRES H R, QUEIRÓS S, MORAIS P, et al. Kidney segmentation in ultrasound, magnetic resonance and computed tomography images: A systematic review[J]. Computer Methods and Programs in Biomedicine, 2018, 157: 49–67. doi: 10.1016/j.cmpb.2018.01.014
    [4] ZHANG Pin, LIANG Yanmei, CHANG Shengjiang, et al. Kidney segmentation in CT sequences using graph cuts based active contours model and contextual continuity[J]. Medical Physics, 2013, 40(8): 081905. doi: 10.1118/1.4812428
    [5] LES T, MARKIEWICZ T, DZIEKIEWICZ M, et al. Adaptive two-way sweeping method to 3D kidney reconstruction[J]. Biomedical Signal Processing and Control, 2021, 67: 102544. doi: 10.1016/j.bspc.2021.102544
    [6] JIN Chao, SHI Fei, XIANG Dehui, et al. 3D fast automatic segmentation of kidney based on modified AAM and random forest[J]. IEEE Transactions on Medical Imaging, 2016, 35(6): 1395–1407. doi: 10.1109/TMI.2015.2512606
    [7] KHALIFA F, SOLIMAN A, TAKIELDEEN A, et al. Kidney segmentation from CT images using a 3D NMF-guided active contour model[C]. The 2016 IEEE 13th International Symposium on Biomedical Imaging, Prague, Czech Republic, 2016: 432–435.
    [8] QAYYUM A, LALANDE A, and MERIAUDEAU F. Automatic segmentation of tumors and affected organs in the abdomen using a 3D hybrid model for computed tomography imaging[J]. Computers in Biology and Medicine, 2020, 127: 104097. doi: 10.1016/j.compbiomed.2020.104097
    [9] 胡敏, 周秀东, 黄宏程, 等. 基于改进U型神经网络的脑出血CT图像分割[J]. 电子与信息学报, 2022, 44(1): 127–137. doi: 10.11999/JEIT200996

    HU Min, ZHOU Xiudong, HUANG Hongcheng, et al. Computed-tomography image segmentation of cerebral hemorrhage based on improved U-shaped Neural Network[J]. Journal of Electronics &Information Technology, 2022, 44(1): 127–137. doi: 10.11999/JEIT200996
    [10] 刘侠, 甘权, 刘晓, 等. 基于超像素的联合能量主动轮廓CT图像分割方法[J]. 光电工程, 2020, 47(1): 190104. doi: 10.12086/oee.2020.190104

    LIU Xia, GAN Quan, LIU Xiao, et al. Joint energy active contour CT image segmentation method based on super-pixel[J]. Opto-Electronic Engineering, 2020, 47(1): 190104. doi: 10.12086/oee.2020.190104
    [11] ÇIÇEK Ö, ABDULKADIR A, LIENKAMP S S, et al. 3D U-Net: Learning dense volumetric segmentation from sparse annotation[C]. The 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, Athens, Greece, 2016: 424–432.
    [12] KANG Li, ZHOU Ziqi, HUANG Jianjun, et al. Renal tumors segmentation in abdomen CT Images using 3D-CNN and ConvLSTM[J]. Biomedical Signal Processing and Control, 2022, 72: 103334. doi: 10.1016/j.bspc.2021.103334
    [13] ZHAN Kun, SHI Jinhui, WANG Haibo, et al. Computational mechanisms of pulse-coupled neural networks: A comprehensive review[J]. Archives of Computational Methods in Engineering, 2017, 24(3): 573–588. doi: 10.1007/s11831-016-9182-3
    [14] BAI Peirui, YANG Kai, MIN Xiaolin, et al. A novel framework for improving Pulse-Coupled Neural Networks with fuzzy connectedness for medical image segmentation[J]. IEEE Access, 2020, 8: 138129–138140. doi: 10.1109/ACCESS.2020.3012160
    [15] 郑瑾, 柳肃, 孙炜. 用于自动识别遥感图像路网信息的改进模糊连接度方法[J]. 电子与信息学报, 2016, 38(2): 413–417. doi: 10.11999/JEIT150563

    ZHENG Jin, LIU Su, and SUN Wei. An improved fuzzy connectedness method to recognize automatically the road network information from remote sensing image[J]. Journal of Electronics &Information Technology, 2016, 38(2): 413–417. doi: 10.11999/JEIT150563
    [16] DE MORAES BRAZ C, MIRANDA P A V, CIESIELSKI K C, et al. Optimum cuts in graphs by general fuzzy connectedness with local band constraints[J]. Journal of Mathematical Imaging and Vision, 2020, 62(5): 659–672. doi: 10.1007/s10851-020-00953-w
    [17] 张睿, 吴薇薇, 周著黄, 等. 基于改进模糊连接度的CT图像肝脏血管三维分割方法[J]. 中国生物医学工程学报, 2019, 38(1): 18–27. doi: 10.3969/j.issn.0258-8021.2019.01.003

    ZHANG Rui, WU Weiwei, ZHOU Zhuhuang, et al. A three-dimensional liver vessel segmentation method for CT images using improved fuzzy connectedness[J]. Chinese Journal of Biomedical Engineering, 2019, 38(1): 18–27. doi: 10.3969/j.issn.0258-8021.2019.01.003
    [18] 李彬, 陈武凡. 基于模糊连接度的多发性硬化症MR图像自动分割算法[J]. 中国生物医学工程学报, 2007, 26(5): 664–668. doi: 10.3969/j.issn.0258-8021.2007.05.005

    LI Bin and CHEN Wufan. Automated segmentation of multiple sclerosis lesions using fuzzy connectedness for MR images[J]. Chinese Journal of Biomedical Engineering, 2007, 26(5): 664–668. doi: 10.3969/j.issn.0258-8021.2007.05.005
    [19] ECKHORN R, REITBOECK H J, ARNDT M, et al. Feature linking via synchronization among distributed assemblies: Simulations of results from cat visual cortex[J]. Neural Computation, 1990, 2(3): 293–307. doi: 10.1162/neco.1990.2.3.293
    [20] CHEN Yuli, PARK S K, MA Yide, et al. A new automatic parameter Setting method of a simplified PCNN for image segmentation[J]. IEEE Transactions on Neural Networks, 2011, 22(6): 880–892. doi: 10.1109/TNN.2011.2128880
    [21] UDUPA J K and SAMARASEKERA S. Fuzzy connectedness and object definition: Theory, algorithms, and applications in image segmentation[J]. Graphical Models and Image Processing, 1996, 58(3): 246–261. doi: 10.1006/gmip.1996.0021
  • 加载中
图(8) / 表(4)
计量
  • 文章访问数:  379
  • HTML全文浏览量:  194
  • PDF下载量:  90
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-09-27
  • 修回日期:  2022-12-08
  • 网络出版日期:  2022-12-09
  • 刊出日期:  2023-06-10

目录

    /

    返回文章
    返回