Dynamic Snake Upsampling Operater and Boundary-Skeleton Weighted Loss for Tubular Structure Segmentation

2025-05-13

AI Summary

Traditional upsampling operators struggle to model elongated, curved topologies of tubular structures (e.g., blood vessels, fissures), leading to segmentation distortion and topological discontinuities in dense prediction tasks such as semantic segmentation and super-resolution. To address this, we propose a Dynamic Snake-like Upsampling (DSU) operator and a Boundary-Skeleton Weighted (BSW) loss function. DSU introduces the first distance-field-guided, adaptive sub-pixel upsampling mechanism along snake-like paths; BSW employs a progressive weighting scheme from skeleton to boundary to jointly optimize overlap accuracy, boundary alignment, and topological continuity. Both modules are lightweight, plug-and-play, and compatible with mainstream backbone architectures. Extensive experiments across multi-domain datasets demonstrate consistent improvements: +3.2% mIoU, −18.7% Hausdorff distance, and +5.9% skeleton F1-score—significantly enhancing both pixel-level accuracy and topological fidelity.

Technology Category

• Computer Vision: Segmentation
• Search and Optimization: Mixed Discrete/Continuous Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Accurate segmentation of tubular topological structures (e.g., fissures and vasculature) is critical in various fields to guarantee dependable downstream quantitative analysis and modeling. However, in dense prediction tasks such as semantic segmentation and super-resolution, conventional upsampling operators cannot accommodate the slenderness of tubular structures and the curvature of morphology. This paper introduces a dynamic snake upsampling operators and a boundary-skeleton weighted loss tailored for topological tubular structures. Specifically, we design a snake upsampling operators based on an adaptive sampling domain, which dynamically adjusts the sampling stride according to the feature map and selects a set of subpixel sampling points along the serpentine path, enabling more accurate subpixel-level feature recovery for tubular structures. Meanwhile, we propose a skeleton-to-boundary increasing weighted loss that trades off main body and boundary weight allocation based on mask class ratio and distance field, preserving main body overlap while enhancing focus on target topological continuity and boundary alignment precision. Experiments across various domain datasets and backbone networks show that this plug-and-play dynamic snake upsampling operator and boundary-skeleton weighted loss boost both pixel-wise segmentation accuracy and topological consistency of results.

Problem

• Accurate segmentation of slender tubular structures in dense prediction tasks
• Dynamic upsampling for subpixel-level feature recovery in serpentine paths
• Balancing main body and boundary focus in topological segmentation loss

Innovation

• Dynamic snake upsampling for tubular structures
• Boundary-skeleton weighted loss function
• Adaptive sampling domain for subpixel recovery