Joint optic disc and cup segmentation based on multi-scale feature analysis and attention pyramid architecture for glaucoma screening

Abstract

Automatic segmentation of optic disc (OD) and optic cup (OC) is an essential task for analysing colour fundus images. In clinical practice, accurate OD and OC segmentation assist ophthalmologists in diagnosing glaucoma. In this paper, we propose a unified convolutional neural network, named ResFPN-Net, which learns the boundary feature and the inner relation between OD and OC for automatic segmentation. The proposed ResFPN-Net is mainly composed of multi-scale feature extractor, multi-scale segmentation transition and attention pyramid architecture. The multi-scale feature extractor achieved the feature encoding of fundus images and captured the boundary representations. The multi-scale segmentation transition is employed to retain the features of different scales. Moreover, an attention pyramid architecture is proposed to learn rich representations and the mutual connection in the OD and OC. To verify the effectiveness of the proposed method, we conducted extensive experiments on two public datasets. On the Drishti-GS database, we achieved a Dice coefficient of 97.59%, 89.87%, the accuracy of 99.21%, 98.77%, and the Averaged Hausdorff distance of 0.099, 0.882 on the OD and OC segmentation, respectively. We achieved a Dice coefficient of 96.41%, 83.91%, the accuracy of 99.30%, 99.24%, and the Averaged Hausdorff distance of 0.166, 1.210 on the RIM-ONE database for OD and OC segmentation, respectively. Comprehensive results show that the proposed method outperforms other competitive OD and OC segmentation methods and appears more adaptable in cross-dataset scenarios. The introduced multi-scale loss function achieved significantly lower training loss and higher accuracy compared with other loss functions. Furthermore, the proposed method is further validated in OC to OD ratio calculation task and achieved the best MAE of 0.0499 and 0.0630 on the Drishti-GS and RIM-ONE datasets, respectively. Finally, we evaluated the effectiveness of the glaucoma screening on Drishti-GS and RIM-ONE datasets, achieving the AUC of 0.8947 and 0.7964. These results proved that the proposed ResFPN-Net is effective in analysing fundus images for glaucoma screening and can be applied in other relative biomedical image segmentation applications.

RMSDSC‐Net: A robust multiscale feature extraction with depthwise separable convolution network for optic disc and cup segmentation

“…The RMSDSC-Net outperforms the recent results by Almubarak et al 32 in DC for OC and OD by (1.5%, 5.5%) on the REFUGE database and Hervella et al 46 in DC for OC and OD by (1.8%, 1.7%) on the DRISHTI-GS database. Finally, we have compared the results with the latest papers from Haider et al, 43 Sun et al, 47 and Hervella et al 48 The proposed approach outperforms the earlier best results by Sun et al 47 on OC DC by around 2.1%. Also, it performs better than the recent best results by Hervella et al 48 and Adnan et al 43 on OC DC on the DRISHTI-GS and REGUFE databases.…”

“…To verify the effectiveness of the proposed approach, we compare our approach with the state‐of‐the‐art approaches, including U‐Net, 24 M‐Net, 25 FC‐DenseNet, 26 pOSAL, 29 Conditional Generative Adversarial Network (GAN), 30 Two‐stage Mask R‐CNN, 32 Team BUCT, 41 Stack‐U‐Net, 44 Wasserstein Generative Adversarial Network (WGAN), 45 Multimodal, 46 ResFPN‐Net, 47 M‐Ada, 48 and SLSR‐Net 43 . Tables 7 and 8 depict the OD and OC segmentation results for different approaches on the DRISHTI‐GS and REFUGE databases, respectively.…”

EARDS: EfficientNet and attention-based residual depth-wise separable convolution for joint OD and OC segmentation

“…In this subsection, we compare the proposed approach with the state-of-the-art approaches, including UNet ( Ronneberger et al, 2015 ), FC-DenseNet ( Al-Bander et al, 2018 ), Yu et al (2019) , WRoIM ( Shah et al, 2019 ), M-Net ( Fu et al, 2018 ), WGAN ( Kadambi et al, 2020 ), pOSAL ( Wang et al, 2019 ), GL-Net ( Jiang et al, 2019 ), CFEA ( Liu et al, 2019 ), Multi-model ( Hervella et al, 2020 ), Two-stage Mask R-CNN ( Almubarak et al, 2020 ), ResFPN-Net ( Sun et al, 2021 ) and M-Ada ( Hervella et al, 2022 ). Table 4 illustrate the OD and OC segmentation results of different approaches on the Drishti-GS and REFUGE datasets, respectively.…”

“…As observed from Table 4 , our approach achieves the best performance in terms of all the evaluation metrics on the two datasets. Finally, the proposed approach is compared with the latest deep learning approaches, i.e., Multi-model ( Hervella et al, 2020 ), Two-stage Mask R-CNN ( Almubarak et al, 2020 ), ResFPN-Net ( Sun et al, 2021 ) and M-Ada ( Hervella et al, 2022 ). According to the results, we can learn that the OD segmentation performance of our approach is slightly lower than ResFPN-Net by 0.0018 (DC) on the Drishti-GS dataset and is inferior to M-Ada by 0.0036 (DC) on the REFUGE dataset.…”

Identifying the Edges of the Optic Cup and the Optic Disc in Glaucoma Patients by Segmentation

“…This architectural improvement leads to a new diffusion-based medical image segmentation method called MedSegDiff-V2 [ 28 ]. Sun et al used ResFPN-Net to learn the boundary features and the inner relation between OD and OC for automatic segmentation [ 29 ]. Xue et al used hybrid level set modeling for disc segmentation [ 30 ].…”