SAR-U-Net: Squeeze-and-excitation block and atrous spatial pyramid pooling based residual U-Net for automatic liver segmentation in Computed Tomography

• Published in: Computer methods and programs in biomedicine Vol. 208; p. 106268
• Main Authors: Wang, Jinke, Lv, Peiqing, Wang, Haiying, Shi, Changfa
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.09.2021
• Subjects: Atrous spatial pyramid pooling; Liver segmentation; Residual structures; Squeeze-and-excitation; U-Net; CT; Liver segmentation; U-Net; Squeeze-and-excitation; Atrous spatial pyramid pooling; Residual structures
• ISSN: 0169-2607; 1872-7565
• DOI: 10.1016/j.cmpb.2021.106268

•A new U-Net based method SAR-U-Net was proposed for liver segmentation.•The attention mechanism is introduced to derive image features in an adaptive manner.•The transition layer and the final output layer of the U-Net decoder is replaced with ASPP to extract richer multi-scale feature.•The standard convolutional layer of U-Net is replaced with the residual block, and attach a batch normalization layer for faster convergence.•Extensive experiments validate the effectiveness of the proposed method. Liver segmentation is an essential prerequisite for liver cancer diagnosis and surgical planning. Traditionally, liver contour is delineated manually by radiologist in a slice-by-slice fashion. However, this process is time-consuming and prone to errors depending on radiologist's experience. In this paper, a modified U-Net based framework is presented, which leverages techniques from Squeeze-and-Excitation (SE) block, Atrous Spatial Pyramid Pooling (ASPP) and residual learning for accurate and robust liver Computed Tomography (CT) segmentation, and the effectiveness of the proposed method was tested on two public datasets LiTS17 and SLiver07. A new network architecture, called SAR-U-Net was designed, which is grounded in the classical U-Net. Firstly, the SE block is introduced to adaptively extract image features after each convolution in the U-Net encoder, while suppressing irrelevant regions, and highlighting features of specific segmentation task; Secondly, the ASPP is employed to replace the transition layer and the output layer, and acquire multi-scale image information via different receptive fields. Thirdly, to alleviate the gradient vanishment problem, the traditional convolution block is replaced with the residual structures, and thus prompt the network to gain accuracy from considerably increased depth. In the LiTS17 database experiment, five popular metrics were used for evaluation, including Dice coefficient, VOE, RVD, ASD and MSD. Compared with other closely related models, the proposed method achieved the highest accuracy. In addition, in the experiment of the SLiver07 dataset, compared with other closely related models, the proposed method achieved the highest segmentation accuracy except for the RVD. An improved U-Net network combining SE, ASPP, and residual structures is developed for automatic liver segmentation from CT images. This new model shows a great improvement on the accuracy compared to other closely related models, and its robustness to challenging problems, including small liver regions, discontinuous liver regions, and fuzzy liver boundaries, is also well demonstrated and validated.