基于全卷积网络的乳腺肿瘤动态增强磁共振图像分割

邱玥,聂生东,魏珑

Segmentation of Breast Tumors Based on Fully Convolutional Network and Dynamic Contrast Enhanced Magnetic Resonance Image

Yue QIU,Sheng-dong NIE,Long WEI

表2 RAU-Net具体细节

Table 2 The details of RAU-Net

结构* 输入 输出 conv 1282*1 1282*64 conv 1282*64 1282*64 max-pooling 1282*64 642*64 RA_block 642* 64 642*64 max-pooling 642*64 322*64 conv 322*64 322*128 conv 322*128 322*128 max-pooling 322*128 162*128 RA_block 162*128 162*128 max-pooling 162*128 82*128 conv 82*256 82*256 conv 82*256 82*256 up-conv 82*256 162*128 merge 162*128, 162*128 162*256 RA_block 162*256 162*256 up-conv 162*256 322*128 merge 322*128, 322*128 322*256 conv 322*256 322*128 conv 322*128 322*128 up-conv 322*128 642*64 merge 642*64, 642*64 642*128 RA_block 642*128 642*128 up-conv 642*128 1282*64 merge 1282*64, 1282*64 1282*128 conv 1282*128 1282*64 conv 1282*64 1282*2 conv 1282*2 1282*1