用Mamba模型统一解决牙片中的四类诊断任务，效率高且精度优。
MATHENA: Mamba-based Architectural Tooth Hierarchical Estimator and Holistic Evaluation Network for Anatomy
- 基于Mamba的多尺度状态空间模型，实现全局上下文建模与牙齿检测
- 四项任务整体性能达标：牙齿检测mAP@50达93.78%，龋齿分割Dice为90.11%
- 支持端到端联合训练与迁移学习，适合临床辅助诊断系统开发
从全景牙片（OPGs）进行牙科诊断需协同完成牙齿检测、龋齿分割（CarSeg）、异常检测（AD）和牙齿发育分期（DDS）。我们提出MATHENA——一种基于Mamba的统一框架，利用其线性复杂度状态空间模型（SSM）同时处理四项任务。MATHENA包含MATHE模块，采用四向视觉状态空间（VSS）块实现O(N)全局上下文建模，生成每颗牙齿的裁剪图像；再由HENA轻量级Mamba-UNet处理，其三头架构中先训练CarSeg建立共享表征，冻结后用于下游AD微调与DDS分类的线性探测，实现稳定高效学习。我们还构建了PARTHENON基准数据集，涵盖来自十个数据集的15,062个标注样本。MATHENA在牙齿检测上达到93.78% mAP@50，CarSeg Dice为90.11%，AD准确率为88.35%，DDS分类准确率为72.40%。
原文摘要 · Abstract (English)
Dental diagnosis from Orthopantomograms (OPGs) requires coordination of tooth detection, caries segmentation (CarSeg), anomaly detection (AD), and dental developmental staging (DDS). We propose Mamba-based Architectural Tooth Hierarchical Estimator and Holistic Evaluation Network for Anatomy (MATHENA), a unified framework leveraging Mamba's linear-complexity State Space Models (SSM) to address all four tasks. MATHENA integrates MATHE, a multi-resolution SSM-driven detector with four-directional Vision State Space (VSS) blocks for O(N) global context modeling, generating per-tooth crops. These crops are processed by HENA, a lightweight Mamba-UNet with a triple-head architecture and Global Context State Token (GCST). In the triple-head architecture, CarSeg is first trained as an upstream task to establish shared representations, which are then frozen and reused for downstream AD fine-tuning and DDS classification via linear probing, enabling stable, efficient learning. We also curate PARTHENON, a benchmark comprising 15,062 annotated instances from ten datasets. MATHENA achieves 93.78% mAP@50 in tooth detection, 90.11% Dice for CarSeg, 88.35% for AD, and 72.40% ACC for DDS.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。