PolypFlow: Reinforcing Polyp Segmentation with Flow-Driven Dynamics

TL;DR

PolypFlow introduces a flow-based model for polyp segmentation that uses physics-inspired ODE dynamics to improve accuracy, interpretability, and robustness against challenging imaging conditions.

Contribution

The paper presents PolypFlow, a novel flow-matching architecture that models segmentation refinement as an ODE, providing interpretability and boundary-aware robustness.

Findings

Achieves state-of-the-art segmentation performance.

Provides interpretable flow-based refinement visualization.

Enhances robustness to low-contrast and motion artifacts.

Abstract

Accurate polyp segmentation remains challenging due to irregular lesion morphologies, ambiguous boundaries, and heterogeneous imaging conditions. While U-Net variants excel at local feature fusion, they often lack explicit mechanisms to model the dynamic evolution of segmentation confidence under uncertainty. Inspired by the interpretable nature of flow-based models, we present \textbf{PolypFLow}, a flow-matching enhanced architecture that injects physics-inspired optimization dynamics into segmentation refinement. Unlike conventional cascaded networks, our framework solves an ordinary differential equation (ODE) to progressively align coarse initial predictions with ground truth masks through learned velocity fields. This trajectory-based refinement offers two key advantages: 1) Interpretable Optimization: Intermediate flow steps visualize how the model corrects under-segmented regions and sharpens boundaries at each ODE-solver iteration, demystifying the ``black-box" refinement process; 2) Boundary-Aware Robustness: The flow dynamics explicitly model gradient directions along polyp edges, enhancing resilience to low-contrast regions and motion artifacts. Numerous experimental results show that PolypFLow achieves a state-of-the-art while maintaining consistent performance in different lighting scenarios.