LRAE-Unet:轻量级MRI脑肿瘤全自动分割网络
LRAE-Unet:a lightweight network for fully automatic segmentation of brain tumor from MRI
林嘉豪 1王瑜 1肖洪兵 1孙梅1
作者信息
- 1. 北京工商大学人工智能学院,北京 100048
- 折叠
摘要
提出一种轻量级脑肿瘤全自动分割网络,即轻量级残差注意力增强网络(LRAE-Unet).首先采用轻量级残差模块解决网络层数增加时出现的梯度消失和网络退化问题;其次采用轻量级自注意力模块抑制输入图像中的不相关区域,同时突出特定局部区域的显著特征;最后通过增强视野平均池化模块减少特征图的空间,节省计算资源,控制网络过拟合现象.在BraTS 2019数据集的测试结果显示LRAE-Unet在完整肿瘤、肿瘤核心与增强肿瘤区域的Dice相似系数为91.24%、88.64%与88.32%,证明使用LRAE-Unet进行脑瘤分割具有可行性与有效性.
Abstract
A lightweight residual attention enhanced Unet(LRAE-Unet)is designed for the fully automatic brain tumor segmentation.LRAE-Unet uses lightweight residual module to solve the problems of gradient disappearance and network degradation when the network layers increases,lightweight self-attention module to suppress the irrelevant areas and highlight the significant features of specific local areas,and enhanced average pooling module with a larger field of perception to reduce the space of feature map,save computing resources and avoid over-fitting.The experiment on BraTS 2019 dataset shows that the proposed method has a Dice similarity coefficient of 91.24%,88.64%and 88.32%in the segmentations of the whole tumor,tumor core and enhanced tumor,which proves its feasibility and effectiveness for brain tumor segmentation.
关键词
脑肿瘤/LRAE-Unet/轻量级残差模块/轻量级自注意力模块/平均池化模块Key words
brain tumor/LRAE-Unet/lightweight residual module/lightweight self-attention module/average pooling module引用本文复制引用
基金项目
北京市自然科学基金-北京市教委科技重点项目(KZ202110011015)
出版年
2024
NETL
NSTL
维普
万方数据