光学精密工程2023，Vol.31Issue(23)：3482-3489,8.DOI:10.37188/OPE.20233123.3482
融合注意力机制的TransGLnet脉络膜自动分割
Automatic segmentation of choroid by TransGLnet integrating attention mechanism
摘要
Abstract
Addressing the challenge posed by the low contrast between the choroid and sclera in choroid segmentation,this research introduces the TransGLnet choroid automatic segmentation network,employ-ing an attention mechanism.The incorporation of a Global Attention Module(GAM)within the convolu-tional layer involves matrix multiplication between features,establishing nonlinear interactions across mul-tiple features in the global spatial context.This enables the extraction of global features without an exces-sive number of parameters.To explore local features,a Local Attention Module(LAM)is introduced be-tween the convolution layer and Transformer encoder,focusing on a 1/4 feature graph.The movement rule for feature graph elements maintains row position consistency while rearranging elements in the col-umn position from largest to smallest.The integration of these two modules ensures that the network effec-tively considers both global and local features.Experimental results showcase the efficacy of the proposed TransGLnet network with a Dice value of 0.91,accuracy at 0.98,equal crossover ratio of 0.89,F1 value reaching 0.90,and a Hausdorff distance of 6.56.Comparative analysis against existing automatic choroi-dal segmentation methods reveals notable improvements in performance metrics.The network presented in this study demonstrates robustness and stability,rendering it suitable for clinical reference.关键词
医学图像处理/脉络膜自动分割/TransUnet/全局注意力/局部注意力Key words
medical image processing/automatic choroidal segmentation/TransUnet/global attention/local attention分类
计算机与自动化引用本文复制引用
黄文博,屈超凡,燕杨..融合注意力机制的TransGLnet脉络膜自动分割[J].光学精密工程,2023,31(23):3482-3489,8.基金项目
吉林省科技厅科技发展计划与自然科学基金联合基金项目(No.YDZJ202101ZYTS147) （No.YDZJ202101ZYTS147）
