Full-scale Representation Guided Network for Retinal Vessel Segmentation

TL;DR

A Full-Scale Guided Network (FSG-Net) is introduced, where a novel feature representation module using modernized convolution blocks effectively captures full-scale structural information, while a guided convolution block subsequently refines this information.

Abstract

The U-Net architecture and its variants have remained state-of-the-art (SOTA) for retinal vessel segmentation over the past decade. In this study, we introduce a Full-Scale Guided Network (FSG-Net), where a novel feature representation module using modernized convolution blocks effectively captures full-scale structural information, while a guided convolution block subsequently refines this information. Specifically, we introduce an attention-guided filter within the guided convolution block, leveraging its similarity to unsharp masking to enhance fine vascular structures. Passing full-scale information to the attention block facilitates the generation of more contextually relevant attention maps, which are then passed to the attention-guided filter, providing further refinement to the segmentation performance. The structure preceding the guided convolution block can be replaced by any U-Net variant, ensuring flexibility and scalability across various segmentation tasks. For a fair comparison, we re-implemented recent studies available in public repositories to evaluate their scalability and reproducibility. Our experiments demonstrate that, despite its compact architecture, FSG-Net delivers performance competitive with SOTA methods across multiple public datasets. Ablation studies further demonstrate that each proposed component meaningfully contributes to this competitive performance. Our code is available on https://github.com/ZombaSY/FSG-Net-pytorch.

Figures (6)

• Figure 1: F1 scores of compared networks on the DRIVE dataset, measured against the validation dataset comprising zero-padded images of resolution 608$\times$608. Among the considered architectures, FSG-Net-T achieved a superior F1 score compared to competitive models while maintaining a reduced parameter size relative to its counterparts. Additionally, the FSG-Net achieved the highest F1 scores while possessing a median parameter size.
• Figure 2: Network architecture of the proposed FSG-Net. In the feature representation network (left), down-convolution layers concatenate separately extracted features, connected via deep bottleneck structures to up-convolution layers. The red dashed boxes highlight feature merging from multiple scales. The guided convolution block (right) employs Guided Residual Modules (GRM) to refine multi-scale features. $S_i$ indicates each stage, and $y_i$ represents labels, where $y_2$ and $y_3$ are downsampled from the original label $y_1$ by interpolation.
• Figure 3: The evolutionary structure from (a): Depthwise residual block, (b): Inverted residual block to (c): The proposed convolution block.
• Figure 4: Detailed structure of guided residual module (GRM)
• Figure 5: Qualitative comparison of the top-3 performing models on the DRIVE validation set.