结合边界自知识蒸馏的结肠镜息肉图像分割方法  

Colonoscopy polyp image segmentation method with boundary self-knowledge distillation

作　　者：孟祥福[1] 张智超 俞纯林 张霄雁[1] Meng Xiangfu;Zhang Zhichao;Yu Chunlin;Zhang Xiaoyan(School of Electronics and Information Engineering,Liaoning Technical University,Huludao 125000,China)

机构地区：[1]辽宁工程技术大学电子与信息工程学院,葫芦岛125000

出　　处：《中国图象图形学报》2025年第2期589-600,共12页Journal of Image and Graphics

基　　金：国家自然科学基金项目(61772249);辽宁省教育厅科学研究项目(LJ2019QL017,LJKZ0355)。

摘　　要：目的结肠镜技术在结肠息肉的早期检测中至关重要,但其依赖于操作员的专业技能和主观判断,因此存在局限性。现有的结肠息肉图像分割方法通常采用额外层和显式扩展网络结构,导致模型效率较低。此外,由于息肉与其周围粘膜之间的边界不清晰,现有模型对于息肉边界的分割效果并不理想。方法提出了一种端到端的自知识蒸馏框架,专门用于结肠息肉图像分割。该框架将边界分割网络和息肉分割网络整合到一个统一的知识蒸馏框架中,以相互增强两个网络的性能。该框架采用专注于边界分割的模型作为教师网络,将息肉分割模型作为学生网络,两者共享一个特征提取模块,以促进更有效的知识传递。设计了一种反向特征融合结构,通过上采样和矩阵乘法聚合编码器深层特征,并利用反向浅层特征作为辅助信息,从而获得分割掩膜的全局映射。结果通过在CVC-Clinic DB(colonoscopy videos challenge-clinicdatabase)、CVC-Colon DB(colonoscopy videos challenge-colondatabase)、Kvasir以及HAM10000(human against machine with 10000 training images)4个数据集上开展实验,与当前11种先进方法Pra Net(parallel reverse attention network)和Polyp2Former(boundary guided network based on transformer for polyp segmentation)等进行比较,实验结果表明本文模型表现最佳,Dice相似性系数(Dice similarity coefficient,DSC)和平均交并比(mean intersection over union,m Io U)指标分别比现有最优模型提升了0.45%和0.68%。结论本文模型适用于各种尺寸和形状的息肉分割,实现了准确的边界提取,并且具有推广到其他医学图像分割任务的潜力。本文代码可在https://github.com/xiaoxiaotuo/BA-KD下载。Objective Colorectal cancer remains a formidable global health challenge,underscoring the pressing need for early detection strategies to improve treatment outcomes.Among these strategies,colonoscopy stands out as a primary diag⁃nostic tool,relying on the visual acumen of medical professionals to identify potentially cancerous abnormalities,such as polyps,within the colon and rectum.However,the effectiveness of colonoscopy is heavily contingent upon the skill and experience of the operator,leading to variability and limitations in detection rates across different practitioners and settings.In response to these challenges,the integration of artificial intelligence and computer vision techniques has garnered increasing attention as a means to augment the accuracy and efficiency of colorectal cancer screening.Various algorithms have been developed to automatically segment colorectal images,with the overarching goal of precisely delineating polyps from the surrounding tissue.Despite advancements in this domain,many existing models confront inherent inefficiencies and limited effectiveness stemming from their intricate architectures and dependence on manual feature engineering.Method This study proposes a novel end-to-end boundary self-knowledge distillation(BA-KD)framework,which aims to achieve precise polyp segmentation.In contrast to conventional methods,BA-KD seamlessly integrates boundary and polyp segmentation networks into a unified framework,facilitating effective knowledge transfer between the two domains.BA-KD represents a pioneering contribution in this field,aiming to harness the synergistic benefits of both boundary and polyp information for increased segmentation accuracy.The BA-KD framework comprises two interconnected branches:a bound⁃ary segmentation network serving as the teacher branch and a polyp segmentation network acting as the student branch.The inherent challenges associated with delineating polyp boundaries are addressed by introducing a boundary detection opera⁃tor to automatically gen

关 键 词：息肉分割 医学图像处理 深度学习 知识蒸馏 边界分割 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]