A Multi-Scale Feature Fusion Method Based on U-Net for Retinal Vessel Segmentation
Computer-aided automatic segmentation of retinal blood vessels plays an important role in the diagnosis of diseases such as diabetes, glaucoma, and macular degeneration. In this paper, we propose a multi-scale feature fusion retinal vessel segmentation model based on U-Net, named MSFFU-Net. The mode...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-07-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/22/8/811 |
| _version_ | 1797561458279579648 |
|---|---|
| author | Dan Yang Guoru Liu Mengcheng Ren Bin Xu Jiao Wang |
| author_facet | Dan Yang Guoru Liu Mengcheng Ren Bin Xu Jiao Wang |
| author_sort | Dan Yang |
| collection | DOAJ |
| description | Computer-aided automatic segmentation of retinal blood vessels plays an important role in the diagnosis of diseases such as diabetes, glaucoma, and macular degeneration. In this paper, we propose a multi-scale feature fusion retinal vessel segmentation model based on U-Net, named MSFFU-Net. The model introduces the inception structure into the multi-scale feature extraction encoder part, and the max-pooling index is applied during the upsampling process in the feature fusion decoder of an improved network. The skip layer connection is used to transfer each set of feature maps generated on the encoder path to the corresponding feature maps on the decoder path. Moreover, a cost-sensitive loss function based on the Dice coefficient and cross-entropy is designed. Four transformations—rotating, mirroring, shifting and cropping—are used as data augmentation strategies, and the CLAHE algorithm is applied to image preprocessing. The proposed framework is tested and trained on DRIVE and STARE, and sensitivity (Sen), specificity (Spe), accuracy (Acc), and area under curve (AUC) are adopted as the evaluation metrics. Detailed comparisons with U-Net model, at last, it verifies the effectiveness and robustness of the proposed model. The Sen of 0.7762 and 0.7721, Spe of 0.9835 and 0.9885, Acc of 0.9694 and 0.9537 and AUC value of 0.9790 and 0.9680 were achieved on DRIVE and STARE databases, respectively. Results are also compared to other state-of-the-art methods, demonstrating that the performance of the proposed method is superior to that of other methods and showing its competitive results. |
| first_indexed | 2024-03-10T18:15:13Z |
| format | Article |
| id | doaj.art-a1a195eb584c4972a54a1ca5d3df1bf2 |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-03-10T18:15:13Z |
| publishDate | 2020-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-a1a195eb584c4972a54a1ca5d3df1bf22023-11-20T07:48:52ZengMDPI AGEntropy1099-43002020-07-0122881110.3390/e22080811A Multi-Scale Feature Fusion Method Based on U-Net for Retinal Vessel SegmentationDan Yang0Guoru Liu1Mengcheng Ren2Bin Xu3Jiao Wang4Key Laboratory of Data Analytics and Optimization for Smart Industry (Northeastern University), Ministry of Education, Shenyang 110819, ChinaCollege of Information Science and Engineering, Northeastern University, Shenyang 110819, ChinaCollege of Information Science and Engineering, Northeastern University, Shenyang 110819, ChinaSchool of Computer Science and Engineering, Northeastern University, Shenyang 110819, ChinaCollege of Information Science and Engineering, Northeastern University, Shenyang 110819, ChinaComputer-aided automatic segmentation of retinal blood vessels plays an important role in the diagnosis of diseases such as diabetes, glaucoma, and macular degeneration. In this paper, we propose a multi-scale feature fusion retinal vessel segmentation model based on U-Net, named MSFFU-Net. The model introduces the inception structure into the multi-scale feature extraction encoder part, and the max-pooling index is applied during the upsampling process in the feature fusion decoder of an improved network. The skip layer connection is used to transfer each set of feature maps generated on the encoder path to the corresponding feature maps on the decoder path. Moreover, a cost-sensitive loss function based on the Dice coefficient and cross-entropy is designed. Four transformations—rotating, mirroring, shifting and cropping—are used as data augmentation strategies, and the CLAHE algorithm is applied to image preprocessing. The proposed framework is tested and trained on DRIVE and STARE, and sensitivity (Sen), specificity (Spe), accuracy (Acc), and area under curve (AUC) are adopted as the evaluation metrics. Detailed comparisons with U-Net model, at last, it verifies the effectiveness and robustness of the proposed model. The Sen of 0.7762 and 0.7721, Spe of 0.9835 and 0.9885, Acc of 0.9694 and 0.9537 and AUC value of 0.9790 and 0.9680 were achieved on DRIVE and STARE databases, respectively. Results are also compared to other state-of-the-art methods, demonstrating that the performance of the proposed method is superior to that of other methods and showing its competitive results.https://www.mdpi.com/1099-4300/22/8/811multi-scaleretinal vessel segmentationU-Netinception structuremax-pooling index |
| spellingShingle | Dan Yang Guoru Liu Mengcheng Ren Bin Xu Jiao Wang A Multi-Scale Feature Fusion Method Based on U-Net for Retinal Vessel Segmentation Entropy multi-scale retinal vessel segmentation U-Net inception structure max-pooling index |
| title | A Multi-Scale Feature Fusion Method Based on U-Net for Retinal Vessel Segmentation |
| title_full | A Multi-Scale Feature Fusion Method Based on U-Net for Retinal Vessel Segmentation |
| title_fullStr | A Multi-Scale Feature Fusion Method Based on U-Net for Retinal Vessel Segmentation |
| title_full_unstemmed | A Multi-Scale Feature Fusion Method Based on U-Net for Retinal Vessel Segmentation |
| title_short | A Multi-Scale Feature Fusion Method Based on U-Net for Retinal Vessel Segmentation |
| title_sort | multi scale feature fusion method based on u net for retinal vessel segmentation |
| topic | multi-scale retinal vessel segmentation U-Net inception structure max-pooling index |
| url | https://www.mdpi.com/1099-4300/22/8/811 |
| work_keys_str_mv | AT danyang amultiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation AT guoruliu amultiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation AT mengchengren amultiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation AT binxu amultiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation AT jiaowang amultiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation AT danyang multiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation AT guoruliu multiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation AT mengchengren multiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation AT binxu multiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation AT jiaowang multiscalefeaturefusionmethodbasedonunetforretinalvesselsegmentation |