坐标增强与多源采样的脑肿瘤图像分割  

Coordinate enhancement and multi-source sampling for brain tumor image segmentation

作　　者：蒋占军[1] 李洋 廉敬[1] 苗新法[1] JIANG Zhanjun;LI Yang;LIAN Jing;MIAO Xinfa(School of Electronic and Information Engineering,Lanzhou Jiaotong University,Lanzhou Gansu 730070,China)

机构地区：[1]兰州交通大学电子与信息工程学院,兰州730070

出　　处：《计算机应用》2025年第3期996-1002,共7页journal of Computer Applications

基　　金：国家自然科学基金资助项目(62061023);甘肃省杰出青年基金资助项目(21JR7RA345)。

摘　　要：针对脑肿瘤图像分割模型对肿瘤区域关注度不够及易丢失空间上下文信息,导致对肿瘤区域分割效果不佳的问题,提出一种融合坐标增强学习机制(CEL)与多源采样的TransUNet脑肿瘤分割网络。首先,提出一种CEL,结合ResNetv2作为模型的浅层特征提取网络,增加对脑肿瘤区域的关注度;其次,设计深层混合采样特征提取器,并利用可变形注意力与自注意力机制对脑肿瘤的全局与局部信息进行多源采样;最后,在编码器与解码器之间设计交互层级融合(ILF)模块,从而在实现深层与浅层特征信息交互的同时减少参数的计算量。在BraTS2018和BraTS2019数据集上的实验结果表明:相较于基准TransUNet,所提模型的平均相似性系数(mDice)、平均交并比(mIoU)、平均精度均值(mAP)和平均召回率(mRecall)分别提高4.84、7.21、3.83和3.15个百分点,模型大小降低了16.9 MB。To address the issues of insufficient focus on tumor regions and the loss of spatial contextual information in brain tumor image segmentation models,which affect the accuracy of tumor segmentation,a TransUNet-based brain tumor segmentation network integrating Coordinate Enhanced Learning mechanism(CEL)and multi-source sampling was proposed.Firstly,a CEL was proposed,and ResNetv2 was combined as shallow feature extraction network of the model,so as to enhance attention to brain tumor regions.Secondly,a deep blended sampling feature extractor was designed,and deformable attention and self-attention mechanisms were used to perform multi-source sampling on both global and local information of brain tumors.Finally,an Interactive Level Fusion(ILF)module was designed between the encoder and the decoder,thereby realizing interaction between deep and shallow feature information while minimizing parameter computational cost.Experimental results on BraTS2018 and BraTS2019 datasets indicate that compared to the benchmark TransUNet,the proposed model has the mean Dice coefficient(mDice),the mean Intersection over Union(mIoU),the mean Average Precision(mAP)and the mean Recall(mRecall)improved by 4.84,7.21,3.83,3.15 percentage points,respectively,and the model size reduced by 16.9 MB.

关 键 词：图像分割 多模态信息 坐标增强学习机制 混合采样 交互层级融合模块 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] R445.2[自动化与计算机技术—计算机科学与技术]