Dual Attention Multiscale Network for Vessel Segmentation in Fundus Photography
Automatic vessel structure segmentation is essential for an automatic disease diagnosis system. The task is challenging due to vessels’ different shapes and sizes across populations. This paper proposes a multiscale network with dual attention to segment various retinal blood vessels. The network in...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-10-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/10/19/3687 |
| _version_ | 1797478087379648512 |
|---|---|
| author | Pengshuai Yin Yupeng Fang Qilin Wan |
| author_facet | Pengshuai Yin Yupeng Fang Qilin Wan |
| author_sort | Pengshuai Yin |
| collection | DOAJ |
| description | Automatic vessel structure segmentation is essential for an automatic disease diagnosis system. The task is challenging due to vessels’ different shapes and sizes across populations. This paper proposes a multiscale network with dual attention to segment various retinal blood vessels. The network injects a spatial attention module and channel attention module on a feature map, whose size is one-eighth of the input size. The network also uses multiscale input to receive multi-level information, and the network uses the multiscale output to gain more supervision. The proposed method is tested on two publicly available datasets: DRIVE and CHASEDB1. The accuracy, AUC, sensitivity, and specificity on the DRIVE dataset are 0.9615, 0.9866, 0.7709, and 0.9847, respectively. On the CHASEDB1 dataset, the metrics are 0.9800, 0.9892, 0.8215, and 0.9877, respectively. The ablative study further shows effectiveness for each part of the network. Multiscale and dual attention mechanism both improve performance. The proposed architecture is simple and effective. The inference time is 12 ms on a GPU and has potential for real-world applications. The code will be made publicly available. |
| first_indexed | 2024-03-09T21:28:02Z |
| format | Article |
| id | doaj.art-e31a97fedddf4d03b7c67f74d9f08ab9 |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-09T21:28:02Z |
| publishDate | 2022-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-e31a97fedddf4d03b7c67f74d9f08ab92023-11-23T21:05:42ZengMDPI AGMathematics2227-73902022-10-011019368710.3390/math10193687Dual Attention Multiscale Network for Vessel Segmentation in Fundus PhotographyPengshuai Yin0Yupeng Fang1Qilin Wan2School of Future Technology, South China University of Technology, Guangzhou 510641, ChinaSchool of Software Engineering, South China University of Technology, Guangzhou 510006, ChinaGuangdong-Hong Kong-Macao Greater Bay Area Weather Research Center for Monitoring Warning and Forecasting, Shenzhen 518000, ChinaAutomatic vessel structure segmentation is essential for an automatic disease diagnosis system. The task is challenging due to vessels’ different shapes and sizes across populations. This paper proposes a multiscale network with dual attention to segment various retinal blood vessels. The network injects a spatial attention module and channel attention module on a feature map, whose size is one-eighth of the input size. The network also uses multiscale input to receive multi-level information, and the network uses the multiscale output to gain more supervision. The proposed method is tested on two publicly available datasets: DRIVE and CHASEDB1. The accuracy, AUC, sensitivity, and specificity on the DRIVE dataset are 0.9615, 0.9866, 0.7709, and 0.9847, respectively. On the CHASEDB1 dataset, the metrics are 0.9800, 0.9892, 0.8215, and 0.9877, respectively. The ablative study further shows effectiveness for each part of the network. Multiscale and dual attention mechanism both improve performance. The proposed architecture is simple and effective. The inference time is 12 ms on a GPU and has potential for real-world applications. The code will be made publicly available.https://www.mdpi.com/2227-7390/10/19/3687vessel segmentationMmedical image analysisdeep learning |
| spellingShingle | Pengshuai Yin Yupeng Fang Qilin Wan Dual Attention Multiscale Network for Vessel Segmentation in Fundus Photography Mathematics vessel segmentation Mmedical image analysis deep learning |
| title | Dual Attention Multiscale Network for Vessel Segmentation in Fundus Photography |
| title_full | Dual Attention Multiscale Network for Vessel Segmentation in Fundus Photography |
| title_fullStr | Dual Attention Multiscale Network for Vessel Segmentation in Fundus Photography |
| title_full_unstemmed | Dual Attention Multiscale Network for Vessel Segmentation in Fundus Photography |
| title_short | Dual Attention Multiscale Network for Vessel Segmentation in Fundus Photography |
| title_sort | dual attention multiscale network for vessel segmentation in fundus photography |
| topic | vessel segmentation Mmedical image analysis deep learning |
| url | https://www.mdpi.com/2227-7390/10/19/3687 |
| work_keys_str_mv | AT pengshuaiyin dualattentionmultiscalenetworkforvesselsegmentationinfundusphotography AT yupengfang dualattentionmultiscalenetworkforvesselsegmentationinfundusphotography AT qilinwan dualattentionmultiscalenetworkforvesselsegmentationinfundusphotography |