发表机构
The University of Auckland; South China University of Technology（奥克兰大学; 华南理工大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出CGDD-Net，通过上下文引导的动态细节建模（CSDE、SAMG、DCDF）实现高效视网膜血管分割，在四个数据集上取得优异性能，仅需196万参数。
AI 中文摘要
准确的视网膜血管分割需要能够捕捉血管几何形态的特征，同时保留用于精细尺度重建的信息。我们提出了CGDD-Net，一种上下文引导的动态细节建模网络，它将自适应特征提取连接到共享解码器路径。上下文引导的尺度自适应可变形编码（CSDE）结合了固定网格卷积与可变形局部注意力，以在多个空间范围内捕捉血管模式。空间自适应多核门控（SAMG）在每个位置选择感受野响应。动态跨尺度细节融合（DCDF）对齐门控中间特征并将其压缩为八通道表示，该表示与选定的编码器跳跃连接一起在三个解码器分辨率下重用。这种设计在解码前整合中间信息，而不是通过单独的直接跳跃传递每个中间阶段特征。在DRIVE、CHASE_DB1、STARE和HRF上，CGDD-Net分别实现了0.9824、0.9938、0.9895和0.9874的AUC值，以及0.8323、0.8102、0.8510和0.8157的F1分数。完整模型包含196万个可训练参数。在累积消融实验中，完整模型在DRIVE、CHASE_DB1和STARE上相比内部基线将F1分别提高了2.48、0.61和3.49个百分点。十二个定向跨数据集实验进一步表征了无目标域适应的迁移性能。结果支持共享中间细节传递作为视网膜血管分割的一种有效且参数紧凑的架构。代码可在 \u0075\u0072\u006c\u007b\u0074\u0068\u0069\u0073\u0020\u0068\u0074\u0074\u0070\u0073\u0020\u0055\u0052\u004c\u007d 获取。
英文摘要
Accurate retinal vessel segmentation requires features that capture vascular geometry while preserving information for fine-scale reconstruction. We propose CGDD-Net, a context-guided dynamic detail modeling network that connects adaptive feature extraction to a shared decoder pathway. Context-Guided Scale-Adaptive Deformable Encoding (CSDE) combines fixed-grid convolution with deformable local attention to capture vascular patterns at multiple spatial extents. Spatially Adaptive Multi-Kernel Gating (SAMG) selects receptive-field responses at each location. Dynamic Cross-Scale Detail Fusion (DCDF) aligns the gated intermediate features and compresses them into an eight-channel representation, which is reused at three decoder resolutions together with selected encoder skips. This design consolidates intermediate information before decoding instead of transferring each middle-stage feature through a separate direct skip. On DRIVE, CHASE\_DB1, STARE, and HRF, CGDD-Net achieves AUC values of 0.9824, 0.9938, 0.9895, and 0.9874, with F1 scores of 0.8323, 0.8102, 0.8510, and 0.8157, respectively. The complete model contains 1.96 million trainable parameters. In cumulative ablations, the full model improves F1 over the internal baseline by 2.48, 0.61, and 3.49 percentage points on DRIVE, CHASE\_DB1, and STARE. Twelve directed cross-dataset experiments further characterize transfer without target-domain adaptation. The results support shared intermediate detail delivery as an effective, parameter-compact architecture for retinal vessel segmentation. Code is available at \url{https://github.com/lixincheng-xcl/CGDD-Net}.
