An improved contrastive learning network for semi-supervised multi-structure segmentation in echocardiography
Cardiac diseases have high mortality rates and are a significant threat to human health. Echocardiography is a commonly used imaging technique to diagnose cardiac diseases because of its portability, non-invasiveness and low cost. Precise segmentation of basic cardiac structures is crucial for cardi...
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2023-09-01
|
| Series: | Frontiers in Cardiovascular Medicine |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fcvm.2023.1266260/full |
| _version_ | 1827809034833494016 |
|---|---|
| author | Ziyu Guo Yuting Zhang Zishan Qiu Suyu Dong Shan He Huan Gao Jinao Zhang Yingtao Chen Bingtao He Zhe Kong Zhaowen Qiu Yan Li Caijuan Li |
| author_facet | Ziyu Guo Yuting Zhang Zishan Qiu Suyu Dong Shan He Huan Gao Jinao Zhang Yingtao Chen Bingtao He Zhe Kong Zhaowen Qiu Yan Li Caijuan Li |
| author_sort | Ziyu Guo |
| collection | DOAJ |
| description | Cardiac diseases have high mortality rates and are a significant threat to human health. Echocardiography is a commonly used imaging technique to diagnose cardiac diseases because of its portability, non-invasiveness and low cost. Precise segmentation of basic cardiac structures is crucial for cardiologists to efficiently diagnose cardiac diseases, but this task is challenging due to several reasons, such as: (1) low image contrast, (2) incomplete structures of cardiac, and (3) unclear border between the ventricle and the atrium in some echocardiographic images. In this paper, we applied contrastive learning strategy and proposed a semi-supervised method for echocardiographic images segmentation. This proposed method solved the above challenges effectively and made use of unlabeled data to achieve a great performance, which could help doctors improve the accuracy of CVD diagnosis and screening. We evaluated this method on a public dataset (CAMUS), achieving mean Dice Similarity Coefficient (DSC) of 0.898, 0.911, 0.916 with 1/4, 1/2 and full labeled data on two-chamber (2CH) echocardiography images, and of 0.903, 0.921, 0.928 with 1/4, 1/2 and full labeled data on four-chamber (4CH) echocardiography images. Compared with other existing methods, the proposed method had fewer parameters and better performance. The code and models are available at https://github.com/gpgzy/CL-Cardiac-segmentation. |
| first_indexed | 2024-03-11T22:36:41Z |
| format | Article |
| id | doaj.art-336990fd7d184fd38cd0a956b31f24ad |
| institution | Directory Open Access Journal |
| issn | 2297-055X |
| language | English |
| last_indexed | 2024-03-11T22:36:41Z |
| publishDate | 2023-09-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Cardiovascular Medicine |
| spelling | doaj.art-336990fd7d184fd38cd0a956b31f24ad2023-09-22T12:58:32ZengFrontiers Media S.A.Frontiers in Cardiovascular Medicine2297-055X2023-09-011010.3389/fcvm.2023.12662601266260An improved contrastive learning network for semi-supervised multi-structure segmentation in echocardiographyZiyu Guo0Yuting Zhang1Zishan Qiu2Suyu Dong3Shan He4Huan Gao5Jinao Zhang6Yingtao Chen7Bingtao He8Zhe Kong9Zhaowen Qiu10Yan Li11Caijuan Li12College of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaSchool of Computer Science, University of Birmingham, Birmingham, United KingdomCollege of Art and Science, New York University Shanghai, Shanghai, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaSchool of Computer Science, University of Birmingham, Birmingham, United KingdomCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaDepartment of Medical Ultrasonics, Hongqi Hospital of Mudanjiang Medical University, Mudanjiang, ChinaCardiac diseases have high mortality rates and are a significant threat to human health. Echocardiography is a commonly used imaging technique to diagnose cardiac diseases because of its portability, non-invasiveness and low cost. Precise segmentation of basic cardiac structures is crucial for cardiologists to efficiently diagnose cardiac diseases, but this task is challenging due to several reasons, such as: (1) low image contrast, (2) incomplete structures of cardiac, and (3) unclear border between the ventricle and the atrium in some echocardiographic images. In this paper, we applied contrastive learning strategy and proposed a semi-supervised method for echocardiographic images segmentation. This proposed method solved the above challenges effectively and made use of unlabeled data to achieve a great performance, which could help doctors improve the accuracy of CVD diagnosis and screening. We evaluated this method on a public dataset (CAMUS), achieving mean Dice Similarity Coefficient (DSC) of 0.898, 0.911, 0.916 with 1/4, 1/2 and full labeled data on two-chamber (2CH) echocardiography images, and of 0.903, 0.921, 0.928 with 1/4, 1/2 and full labeled data on four-chamber (4CH) echocardiography images. Compared with other existing methods, the proposed method had fewer parameters and better performance. The code and models are available at https://github.com/gpgzy/CL-Cardiac-segmentation.https://www.frontiersin.org/articles/10.3389/fcvm.2023.1266260/fullechocardiographydeep learningsemi-supervised learningimages semantic segmentationcontrastive learning |
| spellingShingle | Ziyu Guo Yuting Zhang Zishan Qiu Suyu Dong Shan He Huan Gao Jinao Zhang Yingtao Chen Bingtao He Zhe Kong Zhaowen Qiu Yan Li Caijuan Li An improved contrastive learning network for semi-supervised multi-structure segmentation in echocardiography Frontiers in Cardiovascular Medicine echocardiography deep learning semi-supervised learning images semantic segmentation contrastive learning |
| title | An improved contrastive learning network for semi-supervised multi-structure segmentation in echocardiography |
| title_full | An improved contrastive learning network for semi-supervised multi-structure segmentation in echocardiography |
| title_fullStr | An improved contrastive learning network for semi-supervised multi-structure segmentation in echocardiography |
| title_full_unstemmed | An improved contrastive learning network for semi-supervised multi-structure segmentation in echocardiography |
| title_short | An improved contrastive learning network for semi-supervised multi-structure segmentation in echocardiography |
| title_sort | improved contrastive learning network for semi supervised multi structure segmentation in echocardiography |
| topic | echocardiography deep learning semi-supervised learning images semantic segmentation contrastive learning |
| url | https://www.frontiersin.org/articles/10.3389/fcvm.2023.1266260/full |
| work_keys_str_mv | AT ziyuguo animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT yutingzhang animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT zishanqiu animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT suyudong animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT shanhe animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT huangao animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT jinaozhang animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT yingtaochen animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT bingtaohe animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT zhekong animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT zhaowenqiu animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT yanli animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT caijuanli animprovedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT ziyuguo improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT yutingzhang improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT zishanqiu improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT suyudong improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT shanhe improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT huangao improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT jinaozhang improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT yingtaochen improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT bingtaohe improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT zhekong improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT zhaowenqiu improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT yanli improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography AT caijuanli improvedcontrastivelearningnetworkforsemisupervisedmultistructuresegmentationinechocardiography |