计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (11A): 221100180-7.doi: 10.11896/jsjkx.221100180

• 交叉&应用 • 上一篇    下一篇

基于级联动态注意力U-Net的脑肿瘤分割方法

陈柏年1,2, 韩雨童1, 何涛1,2, 刘斌3, 张建新1,2   

  1. 1 大连民族大学计算机科学与工程学院 大连 116600
    2 大连民族大学机器智能与生物计算研究所 大连 116600
    3 大连理工大学中日国际信息与软件学院 大连 116620
  • 发布日期:2023-11-09
  • 通讯作者: 张建新(jxzhang0411@163.com)
  • 作者简介:(chenbn266@163.com)
  • 基金资助:
    国家自然科学基金(61972062);国家民委中青年英才培养计划项目;辽宁省应用基础研究计划项目

Cascade Dynamic Attention U-Net Based Brain Tumor Segmentation

CHEN Bonian1,2, HAN Yutong1, HE Tao1,2, LIU Bin3, ZHANG Jianxin1,2   

  1. 1 School of Computer Science and Engineering,Dalian Minzu University,Dalian,Liaoning 116600,China
    2 Institute of Machine Intelligence and Bio-computing,Dalian Minzu University,Dalian,Liaoning 116600,China
    3 International School of Information Science and Engineering,Dalian University of Technology,Dalian,Liaoning 116620,China
  • Published:2023-11-09
  • About author:CHEN Bonian,born in 1997,postgra-duate,is a member of China Computer Federation.His main research interests include computer vision and medical image analysis.
    ZHANG Jianxin,born in 1981,Ph.D,professor,master supervisor,is a senior member of China Computer Federation.His main research interests include computer vision and intelligent medical data processing.
  • Supported by:
    National Natural Science Foundation of China(61972062),Young and Middle-aged Talents Program of National Civil Affairs Commission and Applied Basic Research Project of Liaoning Province.

PDF (PC)

202

摘要： 脑肿瘤是一种严重威胁人类健康的疾病,脑肿瘤精确分割在临床诊疗中非常重要。由于脑肿瘤形状大小各异、位置不固定和边界模糊等,实现高精度脑肿瘤自动分割仍是一项具有挑战性的任务。近年来,U-Net凭借其简洁的架构和优秀的性能成为解决医学图像分割任务的主流模型,但其也存在局部感受野有限、空间信息丢失和未充分利用上下文信息等问题。为此,提出一种基于动态卷积和非局部注意力机制的级联U-Net新模型(CDAU-Net)用于脑肿瘤分割任务。首先,将两阶段级联三维U-Net作为主体架构,来重建更精细的高分辨率脑肿瘤空间信息;进而,在级联网络横向连接上添加期望最大化注意力,通过提高网络捕获长距离依赖能力来更好利用肿瘤上下文信息;最后,在级联网络中将普通卷积替换为具有局部自适应能力的动态卷积,可进一步增强网络局部特征捕获能力。在公开的BraTS2019-2020数据集上进行了大量实验并与其他代表性方法进行对比,实验结果表明了所提方法在脑肿瘤分割任务上的有效性。其中,在BraTS2019/2020验证集上获得的全部肿瘤、肿瘤核心和增强肿瘤分割Dice值分别为0.897/0.903,0.826/0.828和0.781/0.786,表现出了良好的脑肿瘤分割性能。

关键词: 脑肿瘤分割, U-Net, 级联网络, 动态卷积, 期望最大化注意力

Abstract: Brain tumor is a common brain disease that heavily threatens human health,so accurate brain tumor segmentation is vital for clinic diagnosis and treatment of patients.Due to different shapes and sizes,unstable positions and fuzzy boundaries of brain tumors,it is a challenging task to achieving high-precision automatic brain tumors segmentation.Recently,U-Net has become the mainstream model for medical image segmentation due to its concise architecture and excellent performance.But it also has some problems,such as limited local receptive field,spatial information loss and insufficient use of context information.Therefore,we propose a new cascade U-Net model based on dynamic convolution and non-local attention mechanism,named CDAU-Net.Firstly,a two-stage cascade 3D U-Net architecture is proposed to reconstruct more detailed and high-resolution spatial information of brain tumors.Then,the expectation-maximization attention is added to the skip connection of CDAU-Net,and the tumor context information is better utilized by improving the network’s ability to capture long-distance dependencies.Finally,the normal convolution is replaced by the dynamic convolution with local adaptive ability in CDAU-Net,which can further enhance the local feature capture ability of the network.Extensive experiments are conducted on public dataset BraTS 2019/2020 and compared with other representative methods,and experimental results show that the proposed method is effective in brain tumor segmentation.The CDAU-Net obtained the Dice values of whole tumor,tumor core and enhancing tumor segmentation on BraTS 2019/2020 verification sets are 0.897/0.903,0.826/0.828 and 0.781/0.786,respectively,which achieves good brain tumor segmentation performance.

Key words: Brain tumor segmentation, U-Net, Cascade network, Dynamic convolution, Expectation-maximizing attention

中图分类号: 

  • TP391
[1]MENZE B H,JAKAB A,BAUER S,et al.The multimodal braintumor image segmentation benchmark(BRATS)[J].IEEE Transactions on Medical Imaging,2014,34(10):1993-2024.
[2]BAUER S,WIEST R,NOLTE L P,et al.A survey of MRI-based medical image analysis for brain tumor studies[J].Physics in Medicine & Biology,2013,58(13):R97.
[3]PEREIRA S,PINTO A,ALVES V,et al.Brain tumor segmen-tation using convolutional neural networks in MRI images[J].IEEE Transactions on Medical Imaging,2016,35(5):1240-1251.
[4]LONG J,SHELHAMER E,DARRELL T.Fully convolutional networks for semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2015:3431-3440.
[5]RONNEBERGER O,FISCHER P,BROX T.U-net:Convolu-tional networks for biomedical image segmentation[C]//International Conference on Medical Image Computing and Compu-ter-Assisted Intervention.Cham:Springer,2015:234-241.
[6]ÇICEK Ö,ABDULKADIR A,LIENKAMP S S,et al.3DU-Net:learning dense volumetric segmentation from sparse annotation[C]//International Conference on Medical Image Computing and Computer-assisted Intervention.Cham:Springer,2016:424-432.
[7]LYU C,SHU H.A two-stage cascade model with variational autoencoders and attention gates for MRI brain tumor segmentation[C]//International MICCAI Brainlesion Workshop.Cham:Springer,2020:435-447.
[8]LIU Z,TONG L,CHEN L,et al.CANet:Context aware network for brain glioma segmentation[J].IEEE Transactions on Medical Imaging,2021,40(7):1763-1777.
[9]AKBAR A S,FATICHAH C,SUCIATI N.Single level UNet3D with multipath residual attention block for brain tumor segmentation[J].Journal of King Saud University－Computer and Information Sciences,2022.
[10]ZHANG J,JIANG Z,LIU D,et al.3D asymmetric expectation-maximization attention network for brain tumor segmentation[J].NMR in Biomedicine,2022,35(5):e4657.
[11]MYRONENKO A.3D MRI brain tumor segmentation usingautoencoder regularization[C]//International MICCAI Brainlesion Workshop.Cham:Springer,2018:311-320.
[12]JIANG Z,DING C,LIU M,et al.Two-stage cascaded u-net:1st place solution to brats challenge 2019 segmentation task[C]//International MICCAI Brainlesion Workshop.Cham:Springer,2019:231-241.
[13]BUATOIS T,PUYBAREAU É,TOCHON G,et al.Two stages CNN-based segmentation of gliomas,uncertainty quantification and prediction of overall patient survival[C]//International MICCAI Brainlesion Workshop.Cham:Springer,2019:167-178.
[14]LI X,LUO G,WANG K.Multi-step cascaded networks forbrain tumor segmentation[C]//International MICCAI Brainlesion Workshop.Cham:Springer,2019:163-173.
[15]LIU C,DING W,LI L,et al.Brain tumor segmentation network using attention-based fusion and spatial relationship constraint[C]//International MICCAI Brainlesion Workshop.Cham:Springer,2020:219-229.
[16]YANG S,GUO D,WANG L,et al.Cascaded Coarse-to-FineNeural Network for Brain Tumor Segmentation[C]//International MICCAI Brainlesion Workshop.Cham,Springer:2020:458-469.
[17]LI X,ZHONG Z,WU J,et al.Expectation-maximization attention networks for semantic segmentation[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2019:9167-9176.
[18]CHEN Y,DAI X,LIU M,et al.Dynamic convolution:Attention over convolution kernels[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:11030-11039.
[19]HU J,SHEN L,SUN G.Squeeze-and-excitation networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2018:7132-7141.
[20]BAKAS S,REYES M,JAKAB A,et al.Identifying the best machine learning algorithms for brain tumor segmentation,progression assessment,and overall survival prediction in the BRATS challenge[J].arXiv:1811.02629,2018.
[21]ZHANG X Y,WANG B,AN W C,et al.Glioma Segmentation Network Based on 3DU-Net++ with Fusion Loss Function[J].Computer Science,2021,48(9):187-193.
[22]XIA F,SHAO H,DENG X.Cross-stage deep-learning-basedMRI fused images of human brain tumor segmentation[J].Journal of Image and Graphics,2022,27(3):873-884.
[23]WANG W,CHEN C,DING M,et al.Transbts:Multimodalbrain tumor segmentation using transformer[C]//International Conference on Medical Image Computing and Computer-Assisted Intervention.Cham:Springer,2021:109-119.
[24]JIANG Y,ZHANG Y,LIN X,et al.SwinBTS:A method for 3D multimodal brain tumor segmentation using swin transformer[J].Brain Sciences,2022,12(6):797.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发