Risk-Routed Implicit Boundary Refinement for Robust Ultrasound Image Segmentation

2026-07-23

AI Summary

Ultrasound image segmentation is highly susceptible to speckle noise, low-contrast boundaries, and acoustic shadowing, often resulting in ambiguous contours and unstable predictions. To address these challenges, this work proposes the RIBR framework, which innovatively employs implicit neural representations as a risk-guided residual refinement module. Rather than directly predicting a complete segmentation mask, RIBR focuses on fine-tuning uncertain boundary regions. The approach integrates a risk-aware routing mechanism to govern residual corrections and incorporates geometric and speckle-aware boundary regularization. Implemented within a compact encoder-decoder architecture, RIBR achieves robust boundary recovery. Evaluated across nine ultrasound datasets, it attains state-of-the-art macro-average performance with significantly fewer parameters and substantially reduces segmentation errors across diverse organ boundaries.

Abstract

Medical ultrasound (US) image segmentation faces significant challenges due to speckle noise, low-contrast boundaries, acoustic shadowing, and acquisition variation across operators and clinical centers. Although encoder-decoder and transformer-based networks have achieved strong performance, many methods recover boundary details through dense decoders or larger backbones, which may still produce over-smoothed contours or unstable predictions under external distribution shifts. In this article, we propose Risk-routed Implicit Boundary Refinement (RIBR), a compact segmentation framework that uses implicit neural representation as a risk-routed residual correction rather than an unconstrained full-mask predictor. RIBR combines boundary-refinement implicit residuals, risk-routed residual control, and geometry- and speckle-aware boundary regularization to refine uncertain contours while suppressing non-boundary oscillations. Evaluation on nine US datasets covering lymph nodes, breast lesions, thyroid nodules, and prostate shows that RIBR achieves the best overall macro-average and consistently reduces boundary error across grouped and organ-specific comparisons under a compact parameter budget. These findings suggest that controlled implicit residual learning is a practical strategy for resource-constrained and boundary-sensitive US segmentation. Source code is available at https://github.com/jinggqu/ribr.

Problem

• ultrasound image segmentation
• speckle noise
• low-contrast boundaries
• acoustic shadowing
• distribution shift

Innovation

• implicit neural representation
• risk-routed residual
• boundary refinement
• ultrasound segmentation
• speckle-aware regularization