Unified Medical Image Segmentation with State Space Modeling Snake

2025-07-16

AI Summary

Unified medical image segmentation (UMIS) faces challenges due to multi-scale organ heterogeneity, complex anatomical morphology, and feature conflicts; conventional pixel-wise methods struggle to model anatomical topology and inter-organ relationships. To address this, we propose Mamba Snake—a novel framework that constructs a hierarchical state-space atlas integrating a snake-shaped visual state-space module (MEB) with energy-map-guided shape priors, enabling joint optimization of macro-scale topological modeling and micro-scale contour evolution. We further introduce a dual-classification synergy mechanism and a detection-segmentation joint training strategy to strengthen multi-organ relational modeling and improve segmentation of small structures. Evaluated on five clinical datasets, Mamba Snake achieves a 3% absolute Dice improvement (relative gain), significantly enhancing segmentation consistency and fidelity of fine anatomical structures. This work establishes a new paradigm for UMIS that jointly incorporates geometric awareness and long-range contextual modeling.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• 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: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Unified Medical Image Segmentation (UMIS) is critical for comprehensive anatomical assessment but faces challenges due to multi-scale structural heterogeneity. Conventional pixel-based approaches, lacking object-level anatomical insight and inter-organ relational modeling, struggle with morphological complexity and feature conflicts, limiting their efficacy in UMIS. We propose Mamba Snake, a novel deep snake framework enhanced by state space modeling for UMIS. Mamba Snake frames multi-contour evolution as a hierarchical state space atlas, effectively modeling macroscopic inter-organ topological relationships and microscopic contour refinements. We introduce a snake-specific vision state space module, the Mamba Evolution Block (MEB), which leverages effective spatiotemporal information aggregation for adaptive refinement of complex morphologies. Energy map shape priors further ensure robust long-range contour evolution in heterogeneous data. Additionally, a dual-classification synergy mechanism is incorporated to concurrently optimize detection and segmentation, mitigating under-segmentation of microstructures in UMIS. Extensive evaluations across five clinical datasets reveal Mamba Snake's superior performance, with an average Dice improvement of 3% over state-of-the-art methods.

Problem

• Addresses multi-scale structural heterogeneity in medical image segmentation
• Overcomes limitations of pixel-based approaches lacking anatomical insight
• Improves segmentation of microstructures and complex morphologies in UMIS

Innovation

• State space modeling for multi-contour evolution
• Mamba Evolution Block for spatiotemporal aggregation
• Dual-classification synergy for detection and segmentation