A Novel Lightweight Swin-Unet Network for Semantic Segmentation of COVID-19 Lesion in CT Images
The Corona Virus Disease 2019 (COVID-19) is highly infectious, has been spread worldwide, caused a global pandemic, and seriously endangered human health and life. The most effective methods for halting and stopping the transmission of the Corona Virus include early detection, quarantine, and succes...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9999672/ |
| _version_ | 1828070262258532352 |
|---|---|
| author | Zhi-Jun Gao Yi He Yi Li |
| author_facet | Zhi-Jun Gao Yi He Yi Li |
| author_sort | Zhi-Jun Gao |
| collection | DOAJ |
| description | The Corona Virus Disease 2019 (COVID-19) is highly infectious, has been spread worldwide, caused a global pandemic, and seriously endangered human health and life. The most effective methods for halting and stopping the transmission of the Corona Virus include early detection, quarantine, and successful treatment. Because it exhibits significant imaging characteristics for COVID-19 lesions in chest computed tomography (CT), it can be used to diagnose COVID-19. Aiming at the inaccuracies of uneven gray distribution, irregular regions, multi-scale, and multi-region segmentation in COVID-19 CT images. This paper proposed a novel Swin-Unet network to improve the accuracy of multi-scale lesion segmentation in COVID-19 CT images. First, in the double-layer Swin Transformer blocks of the Swin-Unet, a residual multi-layer perceptron (ResMLP) module was introduced and replaced the multi-layer perceptron (MLP) module to reduce the loss of features during the transmission process, thereby improving the segmentation precision of multi-scale lesion areas. Second, the uncertain region inpainting module (URIM) was added after Linear Projection, which can refine the uncertain regions in the segmentation features map, thereby improving the segmentation accuracy of different lesion regions. Third, a new loss function DF was designed. It can effectively improve the small target segmentation effect and thus improve the multi-scale segmentation result. Finally, the proposed method was compared to other methods on the public dataset. The Dice, Precision, Recall, and IOU of the proposed method are 0.812, 0.780, 0.848, and 0.683, respectively, which are better than the other models. Moreover, our model has fewer parameters and faster reasoning speed. The proposed method achieves excellent segmentation results for multi-scale and multi-region lesions, and it will be more beneficial in aiding COVID-19 diagnosis and treatment. |
| first_indexed | 2024-04-11T00:35:08Z |
| format | Article |
| id | doaj.art-a0781e46742b47d89010df74e1a1ff9e |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-11T00:35:08Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-a0781e46742b47d89010df74e1a1ff9e2023-01-07T00:00:27ZengIEEEIEEE Access2169-35362023-01-011195096210.1109/ACCESS.2022.32327219999672A Novel Lightweight Swin-Unet Network for Semantic Segmentation of COVID-19 Lesion in CT ImagesZhi-Jun Gao0https://orcid.org/0000-0002-9546-0047Yi He1https://orcid.org/0000-0003-1321-318XYi Li2School of Computer and Information Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaSchool of Computer and Information Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaSchool of Computer and Information Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaThe Corona Virus Disease 2019 (COVID-19) is highly infectious, has been spread worldwide, caused a global pandemic, and seriously endangered human health and life. The most effective methods for halting and stopping the transmission of the Corona Virus include early detection, quarantine, and successful treatment. Because it exhibits significant imaging characteristics for COVID-19 lesions in chest computed tomography (CT), it can be used to diagnose COVID-19. Aiming at the inaccuracies of uneven gray distribution, irregular regions, multi-scale, and multi-region segmentation in COVID-19 CT images. This paper proposed a novel Swin-Unet network to improve the accuracy of multi-scale lesion segmentation in COVID-19 CT images. First, in the double-layer Swin Transformer blocks of the Swin-Unet, a residual multi-layer perceptron (ResMLP) module was introduced and replaced the multi-layer perceptron (MLP) module to reduce the loss of features during the transmission process, thereby improving the segmentation precision of multi-scale lesion areas. Second, the uncertain region inpainting module (URIM) was added after Linear Projection, which can refine the uncertain regions in the segmentation features map, thereby improving the segmentation accuracy of different lesion regions. Third, a new loss function DF was designed. It can effectively improve the small target segmentation effect and thus improve the multi-scale segmentation result. Finally, the proposed method was compared to other methods on the public dataset. The Dice, Precision, Recall, and IOU of the proposed method are 0.812, 0.780, 0.848, and 0.683, respectively, which are better than the other models. Moreover, our model has fewer parameters and faster reasoning speed. The proposed method achieves excellent segmentation results for multi-scale and multi-region lesions, and it will be more beneficial in aiding COVID-19 diagnosis and treatment.https://ieeexplore.ieee.org/document/9999672/COVID-19CT imagesemantic segmentationSwin-UnetResMLP |
| spellingShingle | Zhi-Jun Gao Yi He Yi Li A Novel Lightweight Swin-Unet Network for Semantic Segmentation of COVID-19 Lesion in CT Images IEEE Access COVID-19 CT image semantic segmentation Swin-Unet ResMLP |
| title | A Novel Lightweight Swin-Unet Network for Semantic Segmentation of COVID-19 Lesion in CT Images |
| title_full | A Novel Lightweight Swin-Unet Network for Semantic Segmentation of COVID-19 Lesion in CT Images |
| title_fullStr | A Novel Lightweight Swin-Unet Network for Semantic Segmentation of COVID-19 Lesion in CT Images |
| title_full_unstemmed | A Novel Lightweight Swin-Unet Network for Semantic Segmentation of COVID-19 Lesion in CT Images |
| title_short | A Novel Lightweight Swin-Unet Network for Semantic Segmentation of COVID-19 Lesion in CT Images |
| title_sort | novel lightweight swin unet network for semantic segmentation of covid 19 lesion in ct images |
| topic | COVID-19 CT image semantic segmentation Swin-Unet ResMLP |
| url | https://ieeexplore.ieee.org/document/9999672/ |
| work_keys_str_mv | AT zhijungao anovellightweightswinunetnetworkforsemanticsegmentationofcovid19lesioninctimages AT yihe anovellightweightswinunetnetworkforsemanticsegmentationofcovid19lesioninctimages AT yili anovellightweightswinunetnetworkforsemanticsegmentationofcovid19lesioninctimages AT zhijungao novellightweightswinunetnetworkforsemanticsegmentationofcovid19lesioninctimages AT yihe novellightweightswinunetnetworkforsemanticsegmentationofcovid19lesioninctimages AT yili novellightweightswinunetnetworkforsemanticsegmentationofcovid19lesioninctimages |