arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2507.06363eess.IVcs.CV

Mamba Goes HoME: 分层软专家混合模型用于3D医学图像分割

Mamba Goes HoME: Hierarchical Soft Mixture-of-Experts for 3D Medical Image Segmentation

Faculty of Mathematics, Informatics, and Mechanics, University of Warsaw（华沙大学数学、信息学与力学系）；Faculty of Mathematics and Computer Science, Jagiellonian University（雅盖隆大学数学与计算机科学系）；Institute of AI for Health, Helmholtz Munich（海德堡医学院人工智能与健康研究所）；Faculty of Electronics and Information Technology, Warsaw University of Technology（华沙理工大学电子与信息技术系）；NASK - National Research Institute（国家研究 institute）；Faculty of Radiology, Massachusetts General Hospital（麻省总医院放射学系）；Department of Radiology, Harvard Medical School（哈佛医学院放射学系）

Szymon Płotka, Gizem Mert, Maciej Chrabaszcz, Ewa Szczurek, Arkadiusz Sitek

备注：
Accepted at NeurIPS 2025
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出分层软专家混合模型HoME，通过两级令牌路由提升3D医学图像分割的长上下文建模能力，实现更高效的局部和全局特征提取，从而提升分割性能。

摘要：

近年来，人工智能在医学图像分割领域取得了显著进展。然而，仍存在挑战，包括在多种模态下高效处理3D医学图像以及处理数据变化。本文介绍了一种分层软专家混合模型（HoME），这是一种两级令牌路由层，用于高效长上下文建模，专门设计用于3D医学图像分割。基于Mamba选择性状态空间模型（SSM）主干，HoME通过自适应专家路由增强序列建模。第一层是一个软专家混合层（SMoE），将输入序列分成局部组，将令牌路由到专门的每组专家进行局部特征提取。第二层通过全局SMoE层聚合这些输出，实现跨组信息融合和全局上下文细化。这种分层设计，结合局部专家路由与全局专家细化，增强了通用性和分割性能，超越了在三个最广泛使用的3D医学影像模态和不同数据质量的数据集上的最先进结果。代码可在https://github.com/gmum/MambaHoME上公开获取。

英文摘要：

In recent years, artificial intelligence has significantly advanced medical image segmentation. Nonetheless, challenges remain, including efficient 3D medical image processing across diverse modalities and handling data variability. In this work, we introduce Hierarchical Soft Mixture-of-Experts (HoME), a two-level token-routing layer for efficient long-context modeling, specifically designed for 3D medical image segmentation. Built on the Mamba Selective State Space Model (SSM) backbone, HoME enhances sequential modeling through adaptive expert routing. In the first level, a Soft Mixture-of-Experts (SMoE) layer partitions input sequences into local groups, routing tokens to specialized per-group experts for localized feature extraction. The second level aggregates these outputs through a global SMoE layer, enabling cross-group information fusion and global context refinement. This hierarchical design, combining local expert routing with global expert refinement, enhances generalizability and segmentation performance, surpassing state-of-the-art results across datasets from the three most widely used 3D medical imaging modalities and varying data qualities. The code is publicly available at https://github.com/gmum/MambaHoME.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑