Structure-aware Knowledge-guided Heterogeneous Mamba for Zygomaticomaxillary Suture Assessment

2026-06-15

AI Summary

Assessing the maturity of the zygomaticomaxillary suture (ZMS) is highly challenging due to subtle morphological changes and semantic ambiguity between developmental stages. To address this, this study introduces the first publicly available ZMS dataset comprising 3,790 clinical images and proposes SKMamba, a novel framework featuring a decoupled dual-path Mamba architecture that integrates implicit edge extraction with cross-modal semantic alignment to emulate the diagnostic reasoning of orthodontists. The approach further leverages a large language model to generate anatomical descriptions for interpretability while ensuring morphological evidence remains central to staging decisions. Evaluated on the newly curated dataset, SKMamba significantly outperforms existing methods, achieving high-accuracy automated ZMS staging.

Abstract

The Zygomaticomaxillary Suture is a key circummaxillary structure that connects the zygomatic bone and the maxilla, which serves as a primary site of resistance during maxillary advancement, and its maturation status directly influences the timing and efficacy of orthopedic interventions. However, accurate staging of ZMS maturation remains challenging due to subtle high-frequency transitions in suture lines and the global semantic ambiguity between adjacent stages. To address this, we present the first public ZMS dataset, comprising 3,790 ZMS images covering the entire age range from 4 to 24 years. Based on this dataset, we propose SKMamba, a Structure-aware and Knowledge-guided Mamba-based multi-modal framework for automated ZMS maturation assessment. SKMamba adopts a decoupled dual-path architecture that mimics the hierarchical diagnostic process used by experienced orthodontists. We first introduce an Implicit Edge Extractor (IEE), which leverages structural pre-training to reduce trabecular noise and accentuate sutural boundaries. Complementarily, a Cross-Modal Semantic Alignment (CSA) module is designed to incorporate anatomical descriptions from a large language model (LLM). This module helps align local morphological cues with global semantic descriptions while ensuring that objective morphological evidence remains the primary basis for decisions. Extensive experiments on our ZMS dataset demonstrate that SKMamba achieves state-of-the-art performance compared to existing methods. Code is available at https://github.com/galaxygxq1116/SKMamba.

Problem

• Zygomaticomaxillary Suture
• maturation staging
• suture assessment
• semantic ambiguity
• high-frequency transitions

Innovation

• Structure-aware Mamba
• Implicit Edge Extractor
• Cross-Modal Semantic Alignment
• Zygomaticomaxillary Suture Assessment
• Knowledge-guided Multimodal Learning