DB-KAUNet: An Adaptive Dual Branch Kolmogorov-Arnold UNet for Retinal Vessel Segmentation

2025-12-01

AI Summary

To address the limitations of conventional CNNs in modeling long-range dependencies and complex nonlinear relationships in retinal vessel segmentation, this paper proposes an Adaptive Dual-Branch Kolmogorov–Arnold UNet (KAN-UNet). The architecture introduces a heterogeneous dual-encoder design: one branch employs learnable-activation KANConv layers to replace standard convolutions, while the other integrates a Kolmogorov–Arnold Transformer (KAT) module. Additionally, a geometrically adaptive spatial enhancement module is devised to precisely focus on vessel morphology and suppress background noise. Channel-wise interaction, dynamic attention, and adaptive sampling mechanisms collectively enhance multi-scale feature fusion efficiency. Evaluated on DRIVE, STARE, and CHASE_DB1 benchmarks, KAN-UNet achieves state-of-the-art performance with superior robustness and significantly reduced computational overhead.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Accurate segmentation of retinal vessels is crucial for the clinical diagnosis of numerous ophthalmic and systemic diseases. However, traditional Convolutional Neural Network (CNN) methods exhibit inherent limitations, struggling to capture long-range dependencies and complex nonlinear relationships. To address the above limitations, an Adaptive Dual Branch Kolmogorov-Arnold UNet (DB-KAUNet) is proposed for retinal vessel segmentation. In DB-KAUNet, we design a Heterogeneous Dual-Branch Encoder (HDBE) that features parallel CNN and Transformer pathways. The HDBE strategically interleaves standard CNN and Transformer blocks with novel KANConv and KAT blocks, enabling the model to form a comprehensive feature representation. To optimize feature processing, we integrate several critical components into the HDBE. First, a Cross-Branch Channel Interaction (CCI) module is embedded to facilitate efficient interaction of channel features between the parallel pathways. Second, an attention-based Spatial Feature Enhancement (SFE) module is employed to enhance spatial features and fuse the outputs from both branches. Building upon the SFE module, an advanced Spatial Feature Enhancement with Geometrically Adaptive Fusion (SFE-GAF) module is subsequently developed. In the SFE-GAF module, adaptive sampling is utilized to focus on true vessel morphology precisely. The adaptive process strengthens salient vascular features while significantly reducing background noise and computational overhead. Extensive experiments on the DRIVE, STARE, and CHASE_DB1 datasets validate that DB-KAUNet achieves leading segmentation performance and demonstrates exceptional robustness.

Problem

• Segments retinal vessels for disease diagnosis
• Overcomes CNN limitations in capturing dependencies
• Enhances feature representation with adaptive modules

Innovation

• Dual-branch encoder with CNN and Transformer pathways
• KANConv and KAT blocks for comprehensive feature representation
• Adaptive sampling in SFE-GAF to focus on vessel morphology