SAUNet++: an automatic segmentation model of COVID-19 lesion from CT slices

The coronavirus disease 2019 (COVID-19) epidemic has spread worldwide and the healthcare system is in crisis. Accurate, automated and rapid segmentation of COVID-19 lesion in computed tomography (CT) images can help doctors diagnose and provide prognostic information. However, the variety of lesions and small regions of early lesion complicate their segmentation. To solve these problems, we propose a new SAUNet++ model with squeeze excitation residual (SER) module and atrous spatial pyramid pooling (ASPP) module. The SER module can assign more weights to more important channels and mitigate the problem of gradient disappearance, the ASPP module can obtain context information by atrous convolution using various sampling rates. In addition, the generalized dice loss (GDL) can reduce the correlation between lesion size and dice loss, and is introduced to solve the problem of small regions segmentation. We collected multinational CT scan data from China, Italy and Russia and conducted extensive experiments. In the experiments, SAUNet++ and GDL were compared to advanced segmentation models and popular loss functions, respectively. The experimental results demonstrated that our methods can effectively improve the accuracy of COVID-19 lesion segmentation on the dice similarity coefficient (our: 87.38% VS U-Net++: 86.08%), sensitivity (our: 93.28% VS U-Net++: 89.85%) and hausdorff distance (our: 19.99mm VS U-Net++: 27.69mm), respectively.

[1]  Naimul Mefraz Khan,et al.  A Novel Focal Tversky Loss Function With Improved Attention U-Net for Lesion Segmentation , 2018, 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019).

[2]  S. S. Kumar,et al.  Segmentation of lung from CT using various active contour models , 2019, Biomed. Signal Process. Control..

[3]  M. V. Latte,et al.  Lung Parenchyma Segmentation: Fully Automated and Accurate Approach for Thoracic CT Scan Images , 2018, IETE Journal of Research.

[4]  Yingli Tian,et al.  LGAN: Lung Segmentation in CT Scans Using Generative Adversarial Network , 2019, Comput. Medical Imaging Graph..

[5]  Adetayo Kasim,et al.  The relationship between curvilinear structure enhancement and ridge detection methods , 2020, The Visual Computer.

[6]  Vladlen Koltun,et al.  Multi-Scale Context Aggregation by Dilated Convolutions , 2015, ICLR.

[7]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[8]  Mohamed Abdel-Basset,et al.  FSS-2019-nCov: A deep learning architecture for semi-supervised few-shot segmentation of COVID-19 infection , 2021, Knowledge-Based Systems.

[9]  Zhitao Xiao,et al.  Lung segmentation method with dilated convolution based on VGG-16 network , 2019, Computer assisted surgery.

[10]  Anshu Singla,et al.  Automated segmentation technique with self-driven post-processing for histopathological breast cancer images , 2020, CAAI Trans. Intell. Technol..

[11]  Xiaoqiong Li,et al.  CoSinGAN: Learning COVID-19 Infection Segmentation from a Single Radiological Image , 2020, Diagnostics.

[12]  Gang Sun,et al.  Squeeze-and-Excitation Networks , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[13]  Oscar Camara,et al.  Generalized Overlap Measures for Evaluation and Validation in Medical Image Analysis , 2006, IEEE Transactions on Medical Imaging.

[14]  Narendra D. Londhe,et al.  A deep Residual U-Net convolutional neural network for automated lung segmentation in computed tomography images , 2020 .

[15]  Changming Sun,et al.  RA-UNet: A Hybrid Deep Attention-Aware Network to Extract Liver and Tumor in CT Scans , 2018, Frontiers in Bioengineering and Biotechnology.

[16]  Ling Shao,et al.  Inf-Net: Automatic COVID-19 Lung Infection Segmentation From CT Images , 2020, IEEE Transactions on Medical Imaging.

[17]  V. Radha,et al.  Overview of Computer Aided Detection and Computer Aided Diagnosis Systems for Lung Nodule Detection in Computed Tomography. , 2020, Current medical imaging reviews.

[18]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[19]  S. Hare,et al.  Managing high clinical suspicion COVID-19 inpatients with negative RT-PCR: a pragmatic and limited role for thoracic CT , 2020, Thorax.

[20]  Yu Zhu,et al.  MSD-Net: Multi-Scale Discriminative Network for COVID-19 Lung Infection Segmentation on CT , 2020, IEEE Access.

[21]  Yicheng Fang,et al.  Sensitivity of Chest CT for COVID-19: Comparison to RT-PCR , 2020, Radiology.

[22]  Zhiyong Xu,et al.  A Noise-Robust Framework for Automatic Segmentation of COVID-19 Pneumonia Lesions From CT Images , 2020, IEEE Transactions on Medical Imaging.

[23]  Yap-Peng Tan,et al.  Atrous convolutions spatial pyramid network for crowd counting and density estimation , 2019, Neurocomputing.

[24]  Jun Zhang,et al.  Automatic detection of pulmonary nodules in CT images based on 3D Res-I network , 2020, The Visual Computer.

[25]  Su Ruan,et al.  Automatic COVID‐19 CT segmentation using U‐Net integrated spatial and channel attention mechanism , 2020, Int. J. Imaging Syst. Technol..