|Table of Contents|

Research on Ultrasound Image Segmentation Algorithm forAxillary Lymph Node with Breast Cancer(PDF)

《南京师大学报（自然科学版）》[ISSN:1001-4616/CN:32-1239/N]

Issue:
2021年04期
Page:
122-126
Research Field:
·计算机科学与技术·
Publishing date:

Info

Title:
Research on Ultrasound Image Segmentation Algorithm forAxillary Lymph Node with Breast Cancer
Author(s):
Han Yue1; 2; Zhang Yongshou3; Guo Yiting4; Ban Kaidi2; 5; Cong Jinyu2; 5; Wei Benzheng2; 5
(1.College of Intelligence and Information Engineering,Shandong University of Traditional Chinese Medicine,Jinan 250355,China)(2.Center for Medical Artificial Intelligence,Shandong University of Traditional Chinese Medicine,Qingdao 266112,China)(3.Medical Engineering Department,the 960th Hospital of the PLA,Jinan 250031,China)(4.Department of Radiology,The Second Affiliated Hospital of Shandong University of Traditional Chinese Medicine,Jinan 250001,China)(5.Qingdao Academy of Chinese Medical Sciences,Shandong University of Traditional Chinese Medicine,Qingdao 266112,China)
Keywords:
axillary lymph node; ultrasound image; semantic segmentation; image analysis; deep learning
PACS:
TP391.41 R737.9
DOI:
10.3969/j.issn.1001-4616.2021.04.016
Abstract:
Axillary lymph node is a common metastatic site of breast cancer,and ultrasound image is one of the main ways to detect axillary lymph node metastasis. At present,the segmentation accuracy of axillary lymph node ultrasound image needs to be improved due to its characteristics of much noise and complicate features. In this paper,a new segmentation algorithm based on axillary ultrasound,U-net-MDSC,is proposed. The algorithm automatically segments axillary lymph node by means of decoding and encoding. Aiming at the problems of small ultrasonic image size and low image quality,the unnecessary subsampling structure in the network is reduced. In view of the feature that semantic information is lost a lot in the coding process,a dense jump-join structure is used to extract the image features fully,which can provide information such as the location and size of lymph node. In order to test the effectiveness of the algorithm,the proposed algorithm was verified in the axillary lymph node ultrasound images of 356 patients,and the results showed that intersection over union and Dice coefficient of the proposed algorithm reached 0.838 and 0.903 on the test set.

References:

[1] 袁瑶,周显礼. 乳腺癌超声特征与腋窝淋巴结转移相关性的研究进展[J]. 实用肿瘤学杂志,2020,34(6):576-580.
[2]李智博,周军. 超声对乳腺癌腋窝淋巴结状态评估的应用进展[J]. 实用医学杂志,2020,36(22):3161-3165.
[3]LIU S F,WANG Y,YANG X,et al. Deep learning in medical ultrasound analysis:a review[J]. Engineering,2019,5(2):261-275.
[4]贾海霞. 超声检查在乳腺癌腋窝淋巴结转移中的应用及声像特征研究[J]. 现代医药卫生,2020,36(17):2790-2792.
[5]黄哲兰. 分析超声诊断乳腺癌腋窝淋巴结转移的影像学表现的价值[J]. 影像研究与医学应用,2020,4(21):62-64.
[6]金华,罗伟权,纪宗萍,等. 乳腺癌超声影像组学图像特征Logistic回归方程预测腋窝淋巴结转移风险[J]. 中国超声医学杂志,2021,37(2):139-142.
[7]LIU D M,LAN Y J,ZHANG L,et al. Nomograms for predicting axillary lymph node status reconciled with preoperative breast ultrasound images[J]. Frontiers in oncology,2021,11:567648.
[8]SUN Q C,LIN X N,ZHAO Y S,et al. Deep learning vs. radiomics for predicting axillary lymph node metastasis of breast cancer using ultrasound images:don’t forget the peritumoral region[J]. Frontiers in oncology,2020,10:53.
[9]ZHOU L Q,WU X L,HUANG S Y,et al. Lymph node metastasis prediction from primary breast cancer US images using deep learning[J]. Radiology,2019,294(1):19-28.
[10]ZHENG X Y,YAO Z,HUANG Y N,et al. Deep learning radiomics can predict axillary lymph node status in early-stage breast cancer[J]. Nature communications,2020,11(1):1236.
[11]LIN Y,ZHANG Y Z,CHEN J X,et al. Suggestive annotation:a deep active learning framework for biomedical image segmentation[C]//International Conference on Medical Image Computing and Computer-Assisted Intervention,Quebec City,Canada,2017.Cham:Springer,2017:399-407.
[12]ZHANG Y Z,YING M T C,LIN Y,et al. Coarse-to-fine stacked fully convolutional nets for lymph node segmentation in ultrasound images[C]//IEEE International Conference on Bioinformatics & Biomedicine,Shenzhen,China,2016. Piscataway:IEEE,2016:443-448.
[13]RONNEBERGER O,FISCHER P,BROX T. U-net:convolutional networks for biomedical image segmentation[C]//International Conference on Medical Image Computing and Computer-Assisted Intervention,Munich Germany,2015.Cham:Springer,2015:234-241.

Memo

Memo:
-
Last Update: 2021-12-15