基于级联可分离空洞残差U-Net的肝脏肿瘤分割

doi:10.3969/j.issn.0255-8297.2021.03.004

应用科学学报 ›› 2021, Vol. 39 ›› Issue (3): 378-377.doi: 10.3969/j.issn.0255-8297.2021.03.004

• CCF NCCA 2020专栏 • 上一篇

基于级联可分离空洞残差U-Net的肝脏肿瘤分割

于群¹, 张建新², 魏小鹏^1,3, 张强^1,3

1. 大连大学先进设计与智能计算省部共建教育部重点实验室, 辽宁大连 116622;
2. 大连民族大学计算机科学与工程学院, 辽宁大连 116600;
3. 大连理工大学计算机科学与技术学院, 辽宁大连 116024

收稿日期:2020-08-26 发布日期:2021-06-08
通信作者: 张强，教授，研究方向为智能医学数据分析、生物计算与人工智能。E-mail:zhangq@dlut.edu.cn E-mail:zhangq@dlut.edu.cn
基金资助:
国家重点研发计划基金（No.2018YFC0910500）；国家自然科学基金（No.61972062）；辽宁省重点研发计划项目基金（No.2019JH2/10100030）；辽宁省自然科学基金（No.2019-MS-011）；辽宁省“百千万人才工程”基金资助

Cascaded Separable and Dilated Residual U-Net for Liver Tumor Segmentation

YU Qun¹, ZHANG Jianxin², WEI Xiaopeng^1,3, ZHANG Qiang^1,3

1. Ministry of Education Key Laboratory of Advanced Design and Intelligent Computing, Dalian University, Dalian 116622, Liaoning, China;
2. School of Computer Science and Engineering, Dalian Minzu University, Dalian 116600, Liaoning, China;
3. School of Computer Science and Technology, Dalian University of Technology, Dalian 116024, Liaoning, China

Received:2020-08-26 Published:2021-06-08

摘要/Abstract

摘要： 计算机辅助肝脏肿瘤分割可减少医生工作量，提高手术成功率，因而具有重要的临床诊疗价值。为获得精确的肝脏肿瘤自动分割结果，该文结合医学影像分割领域近年新兴的U-Net模块提出了基于级联可分离空洞残差U-Net（cascaded separable and dilated residualU-Net，CSDResU-Net）的肝脏肿瘤分割方法。CSDResU-Net采用了级联操作，解决了因肿瘤在整幅图像中占比小而造成的肿瘤分割数据不平衡问题；通过在分割网络中整合残差单元、深度可分离卷积和空洞卷积，能够增加卷积核感受野并快速提取更具判别性的肝脏肿瘤图像特征，从而提高肝脏肿瘤分割精度。在国际医学图像计算和计算机辅助干预协会肝脏肿瘤分割数据库上的实验结果表明，CSDResU-Net比基线方法的Dice系数指标提升了1.3%，同时发现空洞率对分割网络的性能表现影响较大。

关键词: U-Net, 残差单元, 空洞卷积, 深度可分离卷积, 肝脏肿瘤分割

Abstract: Computer-aided liver tumor segmentation can effectively reduce the workload of doctors and improve the success rate of surgery, and it has important clinical diagnosis and treatment value. Meanwhile, recently proposed U-Net model has achieved great success in the field of medical image segmentation. To obtain more accurate liver tumor segmentation results, this paper proposed an improved U-net model, i.e., cascaded separable and dilated residual U-Net (CSDResU-Net), for this medical application. CSDResU-Net utilizes cascade operation to solve the problem of unbalanced data in tumor segmentation due to the small proportion of tumors in the whole image. Besides, residual unit, depthwise separable convolution and dilated convolution are integrated into a single network to increase the convolution kernel receptive field, which can quickly extract more discriminative liver image features and lead to the performance improvement of liver tumor segmentation. Experimental results on the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) liver tumor segmentation (LiTS) benchmark dataset show that CSDResU-Net is relative to the baseline. The method improves the performance of the Dice coefficient by 1.3%, and at the same time proves that different void ratios have a greater impact on the performance of the segmentation network.

Key words: U-Net, residual unit, dilated convolution, depthwise separable convolution, liver tumor segmentation

中图分类号:

TP391.41
R73

于群, 张建新, 魏小鹏, 张强. 基于级联可分离空洞残差U-Net的肝脏肿瘤分割[J]. 应用科学学报, 2021, 39(3): 378-377.

YU Qun, ZHANG Jianxin, WEI Xiaopeng, ZHANG Qiang. Cascaded Separable and Dilated Residual U-Net for Liver Tumor Segmentation[J]. Journal of Applied Sciences, 2021, 39(3): 378-377.

参考文献

[1] Mcglynn K A, London W T. The global epidemiology of hepatocellular carcinoma: present and future[J]. Clinics in Liver Disease, 2011, 15(2): 223-243.
[2] 乐美琰，魏千越，邓炜，等. 基于电子计算机断层扫描图像的肝癌病灶自动分割方法研究进展[J]. 生物医学工程学杂志, 2018, 35(3): 481-487, 492.Le M Y, Wei Q Y, Deng W, et al. A review of automatic liver tumor segmentation based on computed tomography[J]. Journal of Biomedical Engineering, 2018, 35(3): 481-487, 492. (in Chinese)
[3] Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[C]//Proceedings of the International Conference on Neural Information Processing Systems, Lake Tahoe, USA, 2012: 1097-1105.
[4] Ke H, Dan C, Li X, et al. Towards brain big data classification: epileptic EEG identification with a lightweight VGGNet on global MIC[J]. IEEE Access, 2018, 6: 14722-14733.
[5] He K M, Zhang X, Ren S Q, et al. Deep residual learning for image recognition[C]//Proceedings of the 29th IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 770-778.
[6] Shelhamer E, Long J, Darrell T. Fully convolutional networks for semantic segmentation[C]//Proceedings of the 28th IEEE Conference on Computer Vision and Pattern Recognition, Boston, USA, 2015: 3431-3440.
[7] Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation[C]//Proceedings of the 18th International Conference on Medical Image Computing and Computer Assisted Intervention, Munich, Germany, 2015: 234-241.
[8] Christ P F, Elshaer M E A, Ettlinger F, et al. Automatic liver and lesion segmentation in CT using cascaded fully convolutional neural networks and 3D conditional random fields[C]//Proceedings of the 19th International Conference on Medical Image Computing and Computer Assisted Intervention, Athens, Greece, 2016: 415-423.
[9] 徐宝泉, 凌彤辉. 基于三维全卷积网络的肝脏和肝癌分割算法研究[J]. 计算机测量与控制, 2019, 27(9): 199-203, 208. Xu B Q, Ling T H. Research on liver and liver tumor segmentation algorithm based on 3D fully convolutional network[J]. Computer Measurement and Control, 2019, 27(9): 199-203, 208. (in Chinese)
[10] Li X M, Chen H, Qi X J, et al. H-DenseUNet: hybrid densely connected UNet for liver and tumor segmentation from CT volumes[J]. IEEE Transactions on Medical Imaging, 2018, 37(12): 2663-2674.
[11] Zhang J P, Xie Y T, Zhang P P, et al. Light-weight hybrid convolutional network for liver tumor segmentation[C]//Proceedings of the International Joint Conference on Artificial Intelligence, Macao, China, 2019: 15-21.
[12] Patrick B, Patrick F C, Eugene V, et al. The liver tumor segmentation benchmark (LiTS)[DB/OL]. Computer Vision and Pattern Recognition, 2019[2020-08-26]. https://arxiv.org/abs/1901.04056.
[13] Kaluva K C, Khened M, Kori A, et al. 2D-densely connected convolution neural networks for automatic liver and tumor segmentation[J/OL]. arXiv: Computer Vision and Pattern Recognition, 2018[2020-08-26]. 1802.02182v1.
[14] Xi X F, Wang L, Sheng V S, et al. Cascade U-ResNets for simultaneous liver and lesion segmentation[J]. IEEE Access, 2020, 8: 68944-68952.

[1]	张育斌, 熊邦书, 欧巧凤, 黄建萍, 陈垚锋. 基于YOLOv3与分水岭的直升机桨叶欠曝光图像圆形标记点检测[J]. 应用科学学报, 2020, 38(6): 906-915.
[2]	尉爽生, 杨忠良, 江旻宇, 黄永峰. 基于神经机器翻译的文本隐写方法[J]. 应用科学学报, 2020, 38(6): 976-985.
[3]	王大千, 崔荣一, 金璟璇. 基于视觉关注模型与多尺度MSER的自然场景文本检测[J]. 应用科学学报, 2020, 38(3): 496-506.
[4]	余淑真, 欧巧凤, 熊邦书, 陈垚锋. 基于散焦图像的大视场立体视觉标定方法[J]. 应用科学学报, 2019, 37(6): 795-805.
[5]	韦秋含, 梁海华, 张新鹏. 基于动态BoW模型的密文JPEG图像检索[J]. 应用科学学报, 2018, 36(4): 628-634.
[6]	史骏鹏, 吴一全. 基于QR分解的NSCT域指纹图像脆弱水印算法[J]. 应用科学学报, 2017, 35(6): 735-744.
[7]	李新伟, 夏秀珍. 基于内容结构图的鲁棒图像哈希[J]. 应用科学学报, 2016, 34(6): 691-701.
[8]	孙物一, 冯国瑞. 基于高维特征正则化的JPEG图像隐写分析[J]. 应用科学学报, 2016, 34(5): 555-563.
[9]	孟天亮, 吴一全, 吴诗婳. 基于引导滤波和NSST的工业CT图像边缘检测[J]. 应用科学学报, 2016, 34(4): 405-416.
[10]	胡昭华, 王冠南, 王珏, 邵晓雯, 卞飞飞. 模板在线聚类的目标跟踪[J]. 应用科学学报, 2016, 34(4): 430-440.
[11]	刘华勇, 谢新平, 李璐, 张大明. 三角Bézier曲线曲面光滑融合的构造[J]. 应用科学学报, 2016, 34(2): 154-162.
[12]	张莉, 叶志伟, 王明威. 基于差分进化的二维熵图像分割[J]. 应用科学学报, 2016, 34(1): 58-66.
[13]	吴诗婳1，吴一全1,2,3,4,5，周建江1，孟天亮1，戴一冕1. 基于Shearlet 变换和Krawtchouk 矩不变量的河流SAR 图像分割[J]. 应用科学学报, 2015, 33(1): 21-31.
[14]	郑毅1,2，郑苹3,4. 结合模糊熵和遗传算法的双阈值图像分割[J]. 应用科学学报, 2014, 32(4): 427-433.
[15]	胡蓓1，隆霞1，胡超1，段盼2，唐若笠1，段其昌1. 改进脉冲耦合神经网络及二维Otsu算法的光伏阵列阴影检测[J]. 应用科学学报, 2013, 31(6): 613-618.

基于级联可分离空洞残差U-Net的肝脏肿瘤分割

Cascaded Separable and Dilated Residual U-Net for Liver Tumor Segmentation

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价