Segmentation network for metastatic lymph nodes of head and neck tumors

被引：0
|
作者：
Zhou T. [1 ,2 ]
Shi D. [1 ,2 ]
Xue J. [3 ]
Peng C. [1 ,2 ]
Dang P. [1 ,2 ]
Zhou Z. [3 ]
机构：
[1] College of Computer Science and Engineering, North Minzu University, Yinchuan
[2] Key Laboratory of Image and Graphics Intelligent Processing of State Ethnic Affairs Commission, North Minzu University, Yinchuan
[3] College of Oral Cavity, Ningxia Medical University, Yinchuan
来源：
Guangxue Jingmi Gongcheng/Optics and Precision Engineering | 2024年 / 32卷 / 09期
关键词：
attention mechanism; head and neck tumors; instance segmen⁃tation; lymph node metastasis; medical image processing;
D O I：
10.37188/OPE.20243209.1420
中图分类号：
学科分类号：
摘要：
Head and neck tumors are prevalent malignant tumors in China, with prognosis significantly in⁃ fluenced by cervical lymph node metastasis. In medical practice, magnetic resonance imaging (MRI) is em⁃ ployed to identify metastatic lymph nodes. However, MRI images often suffer from blurred edges and low contrast between the lesion and surrounding tissue. This paper introduces a segmentation network tailored for metastatic lymph nodes in head and neck tumors. Initially, a cross-layer and cross-field attention mod⁃ ule is developed, integrating features from both deep and shallow layers to enhance the shape representa⁃ tion of metastatic lymph nodes through a self-attention mechanism. This module improves contextual se⁃ mantic understanding across different receptive fields, allowing for pixel-level fusion of shallow and deep feature maps, thereby enhancing the morphological details of metastatic lymphatic nodes. Subsequently, a multi-scale feature fusion module is designed to amalgamate features across various scales in the feature pyramid, enriching the morphological details of the lymph nodes. Furthermore, an enhanced attention pre⁃ diction head module is implemented, combining parallel self-attention and gate channel transformation to accentuate the lesion area and refine its boundaries on the feature map. The network′s effectiveness is con⁃ firmed using a clinical dataset of lymph node metastasis medical images. The performance metrics, AP⁃ det, APseg, ARdet, ARseg, mAPdet, and mAPseg for lymph node metastasis lesion segmentation are 74.88%, 74.12%, 63.11%, 62.28%, 74.64%, and 74.04%, respectively. This network provides pre⁃ cise detection and segmentation of lymph node metastasis lesions, offering significant benefits for lymph node diagnosis. © 2024 Chinese Academy of Sciences. All rights reserved.
引用
收藏
页码：1420 / 1431
页数：11
相关论文
共 21 条
  • [1] PYNNONEN M A., Clinical practice guideline: Evaluation of the neck mass in adults［J］. Otolaryn⁃ gol, Head Neck Surg, 157, 2, (2017)
  • [2] DILLEKAS H, ROGERS M S, STRAUME O., Are 90% of deaths from cancer caused by metasta⁃ ses?［J］, Cancer Medicine, 8, 12, pp. 5574-5576, (2019)
  • [3] LEE S, CHO M, LEE W, Et al., Predicting lymph node metastasis and distant metastasis using differen⁃ tial correlations of miRNAs and their target RNAs in cancer［C］, 2022 IEEE International Conference on Bioinformatics and Biomedicine(BIBM), pp. 457-463, (2022)
  • [4] OYAMA D, ADACHI Y, TSUYUGUCHI N., Magnetic resonance imaging using a magnetoresis⁃ tive sensor with a flux transformer［J］, IEEE Trans⁃ actions on Magnetics, 59, 11, (2023)
  • [5] WANG Q, QIAO K, QIN R X, Et al., Predictive model of breast cancer lymph node metastasis based on deep learning(E-Transformer)［C］, 2021 6th In⁃ ternational Symposium on Computer and Informa⁃ tion Processing Technology(ISCIPT), pp. 168-173, (2021)
  • [6] ZHOU T, DANG P, LU H L, Et al., A transform⁃ er segmentation model for PET/CT images with cross-modal［J/OL］, Journal of Electronics & Infor⁃ mation, pp. 1-9, (2022)
  • [7] GURNEY-CHAMPION O J, KIESELMANN J P, WONG K H, Et al., A convolutional neural net⁃ work for contouring metastatic lymph nodes on diffu⁃ sion-weighted magnetic resonance images for assess⁃ ment of radiotherapy response［J］, Physics and Im⁃ aging in Radiation Oncology, 15, pp. 1-7, (2020)
  • [8] TONG Z X, KANG Y X, YING H, Et al., A novel framework for lymphatic invasion detection based on efficient medical images translation network［C］, 2021 6th International Conference on Intelligent Computing and Signal Processing(ICSP), pp. 669-673
  • [9] LUO Y W, XIN J M, LIU S J, Et al., Lymph node metastasis classification based on semi-supervised multi-view network［C］, 2020 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 675-680, (2020)
  • [10] ZHAO X Y, XIE P Y, WANG M M, Et al., Deep learning-based fully automated detection and seg⁃ mentation of lymph nodes on multiparametric-MRI for rectal cancer:a multicentre study［J］, EBio⁃ Medicine, 56, (2020)
123