[1]蔡体健,蒋嘉豪,刘遵雄,等.基于局部上下文融合的息肉语义分割模型[J].中国医学物理学杂志,2025,42(1):128-134.[doi:DOI:10.3969/j.issn.1005-202X.2025.01.017]
　CAI Tijian,JIANG Jiahao,LIU Zunxiong,et al.Polyp semantic segmentation model based on local context fusion[J].Chinese Journal of Medical Physics,2025,42(1):128-134.[doi:DOI:10.3969/j.issn.1005-202X.2025.01.017]
点击复制

基于局部上下文融合的息肉语义分割模型()
分享到：

《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]

卷:
42
期数:
2025年第1期
页码:
128-134
栏目:
医学人工智能
出版日期:
2025-01-19

文章信息/Info

Title:
Polyp semantic segmentation model based on local context fusion
文章编号:
1005-202X（2025）01-0128-07
作者:
蔡体健; 蒋嘉豪; 刘遵雄; 赵师明; 易晟权
华东交通大学信息与软件工程学院， 江西 南昌 330013
Author(s):
CAI Tijian;  JIANG Jiahao;  LIU Zunxiong;  ZHAO Shiming;  YI Shengquan
School of Information and Software Engineering, East China Jiaotong University, Nanchang 330013, China
关键词:
结直肠癌; 息肉分割; 深度学习; 扩展卷积; 上下文信息
Keywords:
Keywords: colorectal cancer polyp segmentation deep learning dilated convolution context information
分类号:
R318
DOI:
DOI:10.3969/j.issn.1005-202X.2025.01.017
文献标志码:
A
摘要:
为了准确分割息肉，提出了局部上下文融合的分割模型，该模型引入局部上下文注意力机制来过滤掉无关的特征信息，并增强对重要区域的关注。通过多核扩展卷积捕获不同尺度的特征，以提高息肉边界分割的精度。引入金字塔上下文选择模块，利用较浅层编码器特征补偿深层编码器丢失的低级信息，使模型能够适应各种大小的息肉。该模型在Kvasir-SEG、EndoScene和CVC-ClinicDB数据集上分别达到了97.67%、97.19%和99.23%的准确率，平均交并比分别为91.2%、88.31%和94.75%，比现有经典方法有更好的准确性和泛化性，验证了该模型在息肉分割任务中的优越性能。本文模型在息肉分割准确性方面得到提升，为息肉分割提供更为精准的辅助。
Abstract:
Abstract: A local context fusion based segmentation model which uses a local context attention mechanism to filter out irrelevant feature information and enhance the attention to important regions is presented for accurate polyp segmentation. The features at different scales are captured by multi-kernel dilated convolution for improving the accuracy of polyp boundary segmentation. Pyramid context selection module utilizes shallow encoder features to compensate for the low-level information lost by the deeper encoder, enabling the model to adapt to polyps of various sizes. The proposed model achieves accuracies of 97.67%, 97.19% and 99.23% on Kvasir-SEG, EndoScene and CVC-ClinicDB datasets, respectively, with mIoU of 91.20%, 88.31% and 94.75%, respectively, exhibiting higher accuracy and generalizability than the existing classical methods and validating its superior performance in polyp segmentation. The proposed model can improve polyp segmentation accuracy and provide a more accurate aid for polyp segmentation.

相似文献/References:

[1]薛芳沁,陈鸿源,许 超,等.纳米炭示踪在结直肠癌根治术淋巴结检出中的应用[J].中国医学物理学杂志,2015,32(04):490.[doi:10.3969/j.issn.1005-202X.2015.04.008]
　[J].Chinese Journal of Medical Physics,2015,32(1):490.[doi:10.3969/j.issn.1005-202X.2015.04.008]
[2]喻娟,胡春宏. 奥曲肽防治结直肠癌根治术后放疗致腹泻的效果[J].中国医学物理学杂志,2017,34(4):355.[doi:DOI:10.3969/j.issn.1005-202X.2017.04.007]
[3]林盛敏,陈小云.超声引导下腰方肌阻滞联合丙泊酚麻醉在腹腔镜结直肠癌根治术中的应用[J].中国医学物理学杂志,2022,39(5):567.[doi:DOI:10.3969/j.issn.1005-202X.2022.05.008]
　LIN Shengmin,CHEN Xiaoyun.Application of ultrasound-guided quadratus lumborum block combined with propofol anesthesia in laparoscopic radical resection for colorectal cancer[J].Chinese Journal of Medical Physics,2022,39(1):567.[doi:DOI:10.3969/j.issn.1005-202X.2022.05.008]
[4]刘宏滨,顾德.融合Transformer和卷积的结直肠息肉分割算法[J].中国医学物理学杂志,2024,41(3):316.[doi:DOI:10.3969/j.issn.1005-202X.2024.03.008]
　LIU Hongbin,GU De.Colorectal polyp segmentation algorithm integrating Transformer and convolution[J].Chinese Journal of Medical Physics,2024,41(1):316.[doi:DOI:10.3969/j.issn.1005-202X.2024.03.008]

备注/Memo

备注/Memo:
【收稿日期】2024-09-26 【基金项目】国家自然科学基金（62166018）；江西省研究生创新专项资金项目（YC2023-S531）；江西省自然科学基金（20232BAB202055） 【作者简介】蔡体健，博士，副教授，硕士生导师，研究方向：计算机视觉、深度学习、稀疏表示，E-mail: cai2017@ecjtu.edu.cn
更新日期/Last Update: 2025-01-19