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

Abstract

BackgroundGlaucoma is the leading cause of irreversible vision loss. Accurate Optic Disc (OD) and Optic Cup (OC) segmentation is beneficial for glaucoma diagnosis. In recent years, deep learning has achieved remarkable performance in OD and OC segmentation. However, OC segmentation is more challenging than OD segmentation due to its large shape variability and cryptic boundaries that leads to performance degradation when applying the deep learning models to segment OC. Moreover, the OD and OC are segmented independently, or pre-requirement is necessary to extract the OD centered region with pre-processing procedures.MethodsIn this paper, we suggest a one-stage network named EfficientNet and Attention-based Residual Depth-wise Separable Convolution (EARDS) for joint OD and OC segmentation. In EARDS, EfficientNet-b0 is regarded as an encoder to capture more effective boundary representations. To suppress irrelevant regions and highlight features of fine OD and OC regions, Attention Gate (AG) is incorporated into the skip connection. Also, Residual Depth-wise Separable Convolution (RDSC) block is developed to improve the segmentation performance and computational efficiency. Further, a novel decoder network is proposed by combining AG, RDSC block and Batch Normalization (BN) layer, which is utilized to eliminate the vanishing gradient problem and accelerate the convergence speed. Finally, the focal loss and dice loss as a weighted combination is designed to guide the network for accurate OD and OC segmentation.Results and discussionExtensive experimental results on the Drishti-GS and REFUGE datasets indicate that the proposed EARDS outperforms the state-of-the-art approaches. The code is available at https://github.com/M4cheal/EARDS.

Deep learning nomogram for predicting neoadjuvant chemotherapy response in locally advanced gastric cancer patients

“…In comparison to the original EfficientNet, EfficientNet V2 incorporates more efficient architectural designs and optimization strategies, achieving improved performance while maintaining a relatively smaller model size. It combines optimizations in model depth, width, and resolution to provide enhanced accuracy, particularly in scenarios with limited computational resources [37].…”

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

“… where a true positive is represented by TP , a false positive is represented by FP , and a false negative is represented by FN [ 27 ].…”

“…Parkhi et al utilized DeepLabv3 and ensemble models to perform the segmentation of the optic disc and cup [ 26 ]. Zhou et al developed a one-stage network named EfficientNet and Attention-based Residual Depth-Wise Separable Convolution (EARDS) for joint OD and OC segmentation [ 27 ]. Wu et al developed a transformer-based conditional U-Net framework and a new Spectrum-Space Transformer to model the interaction between noise and semantic features.…”

HDR-EfficientNet: A Classification of Hypertensive and Diabetic Retinopathy Using Optimize EfficientNet Architecture

“…The EfficientNet model, an upgraded version of EfficientNet V2-B4, is provided to detect HR and DR eye disorders. A further developed version of EfficientNet is Efficient-NetV2 [31]. The updated EfficientNetV2 model is essentially offered to expand the resources available while retaining a high recall rate.…”

“…Augment Time (S) ACC RBM-DNN [25] Yes 38 79.1% CNN-technique [26] Yes 55 83.5% EfficientNet [31] Yes 35 78.84% ResNet50 [23] Yes 42 73.75% DenseNet-101 [24] Yes 39 93.93% Proposed HDR-EfficientNet Yes 30 98.12% , IDRiD [33], MESSIDOR [34]), and (b) shows the datasets obtained from (Kaggle-Dataset [32], IDRiD [33], MESSIDOR [34], e-ophtha [35], HRF [36], and EYEPACS [37]).…”