多模态信息融合 - arXivDaily 专题

2508.01819 2026-06-19 eess.IV 版本更新 80%

Decoding the Alzheimer's Continuum: Interpretable Multi-Gate Routing for Diagnosis and Transition Prediction

解码阿尔茨海默病连续谱：可解释的多门路由用于诊断与转换预测

Yufeng Jiang, Hexiao Ding, Hongzhao Chen, Jing Lan, Xinzhi Teng, Gerald W. Y. Cheng, Yunlin Mao, Zongxi Li, Haoran Xie, Jung Sun Yoo, Jing Cai

专题命中医学影像融合：多门专家混合架构融合临床先验与MRI

AI总结提出M$^3$AD统一框架，利用可解释多门专家混合架构，基于T1加权sMRI同时实现三分类诊断和阶段转换预测，准确率达95.13%。

Comments Accepted by MICCAI2026

详情

AI中文摘要

阿尔茨海默病（AD）表现为从正常认知（NC）经轻度认知障碍（MCI）到痴呆的连续进展。然而，大多数深度学习方法将此连续谱简化为不连续的分类任务，很大程度上忽略了动态阶段转换。为了解码这一复杂进展，我们提出M$^3$AD，一个统一框架，仅使用T1加权sMRI联合处理三分类诊断和诊断阶段转换预测。M$^3$AD利用可解释的多门专家混合架构，采用专门的路由机制动态捕获诊断特定的病理模式和跨连续谱的共享结构特征。它进一步通过自适应注意力融合整合临床先验（年龄、性别、eTIV）以增强泛化能力。M$^3$AD在原始实验设置下达到95.13%的准确率（MCLNC报告为90.44%），转换预测准确率为94.87%。关键的是，分析多门路由揭示了区分稳定性和进展性MCI的独特专家激活特征，为个体水平的进展风险分层提供了机制基础。代码见：此 https URL。

英文摘要

Alzheimer's disease (AD) manifests as a continuous progression from normal cognition (NC) through mild cognitive impairment (MCI) to dementia. However, most deep learning approaches reduce this continuum to disjointed classification tasks, largely ignoring dynamic stage transitions. To decode this complex progression, we propose M$^3$AD, a unified framework that jointly addresses three-class diagnosis classification and diagnosis stage transition prediction using only T1-weighted sMRI. M$^3$AD leverages an interpretable multi-gate mixture of experts architecture, employing specialized routing mechanisms to dynamically capture both diagnosis-specific pathological patterns and shared structural features across the continuum. It further integrates clinical priors (age, sex, eTIV) via adaptive attention fusion to enhance generalization. M$^3$AD achieves 95.13% accuracy, compared to 90.44% reported by MCLNC under its original experimental setting, and 94.87% for transition prediction. Crucially, analyzing the multi-gate routing reveals distinct expert activation signatures distinguishing stable from progressive MCI, providing a mechanistic basis for individual-level progression risk stratification. Code is available at https://github.com/csyfjiang/M3AD.

URL PDF HTML ☆

赞 0 踩 0

2503.23179 2026-06-19 eess.IV cs.CV 版本更新 80%

OncoReg: Medical Image Registration for Oncological Challenges

OncoReg：面向肿瘤学挑战的医学图像配准

Wiebke Heyer, Yannic Elser, Lennart Berkel, Xinrui Song, Xuanang Xu, Pingkun Yan, Xi Jia, Jinming Duan, Zi Li, Tony C. W. Mok, BoWen LI, Tim Hable, Christian Staackmann, Christoph Großbröhmer, Lasse Hansen, Alessa Hering, Malte M. Sieren, Mattias P. Heinrich

发表机构 * Institute of Medical Informatics, University of Lübeck（吕贝克大学医学信息学研究所）； Institute of Radiology and Nuclear Medicine, University Hospital Schleswig-Holstein（石勒斯维希-霍尔斯坦大学医院放射科和核医学研究所）； Department of Biomedical Engineering and Center for Biotechnology and Interdisciplinary Studies, Rensselaer Polytechnic Institute（伦塞拉塞尔理工学院生物医学工程系和生物技术与跨学科研究中心）； School of Computer Science, University of Birmingham（伯明翰大学计算机科学学院）； Division of Informatics, Imaging and Data Sciences, University of Manchester（曼彻斯特大学信息学、成像和数据科学系）； DAMO Academy, Alibaba Group（阿里集团DAMO学院）； Hangzhou Shengshi Technology Co., Ltd（杭州盛世科技有限公司）； Department of Radiation Oncology, University Hospital Schleswig-Holstein（石勒斯维希-霍尔斯坦大学医院放射肿瘤科）； EchoScout GmbH ； Radboud University Medical Center, Nijmegen（奈密根大学医学中心）； Institute of Interventional Radiology, University Hospital Schleswig-Holstein（石勒斯维希-霍尔斯坦大学医院介入放射科）

专题命中医学影像融合：CBCT与FBCT配准，属于医学影像融合

AI总结提出OncoReg挑战，通过两阶段框架在保护患者隐私的同时开发可泛化的图像配准方法，用于放射治疗中锥束CT与扇束CT的配准，发现特征提取是关键，深度学习和经典方法结合最有效。

Comments 21 pages, 13 figures

详情

AI中文摘要

在现代癌症研究中，由于患者隐私相关的挑战，产生的大量医学数据往往未被充分利用。OncoReg挑战通过一个两阶段框架解决了这一问题，该框架使研究人员能够在确保患者隐私的同时开发和验证图像配准方法，并促进更可泛化的AI模型的发展。第一阶段涉及使用公开可用的数据集，第二阶段则专注于在安全的医院网络内对私有数据集进行模型训练。OncoReg建立在Learn2Reg挑战的基础上，纳入了放射治疗中介入性锥束计算机断层扫描与标准计划扇束CT图像的配准。准确的图像配准在肿瘤学中至关重要，特别是在图像引导放射治疗的动态治疗调整中，需要精确对齐以最小化对健康组织的辐射暴露，同时有效靶向肿瘤。本文详细介绍了OncoReg挑战的方法和数据，并对竞赛参赛作品和结果进行了全面分析。研究发现，特征提取在此配准任务中起着关键作用。从该挑战中涌现的一种新方法展示了其多功能性，而现有方法的表现与新技术相当。深度学习和经典方法在图像配准中仍扮演重要角色，尤其是方法的组合，特别是在特征提取方面，被证明最为有效。

英文摘要

In modern cancer research, the vast volume of medical data generated is often underutilised due to challenges related to patient privacy. The OncoReg Challenge addresses this issue by enabling researchers to develop and validate image registration methods through a two-phase framework that ensures patient privacy while fostering the development of more generalisable AI models. Phase one involves working with a publicly available dataset, while phase two focuses on training models on a private dataset within secure hospital networks. OncoReg builds upon the foundation established by the Learn2Reg Challenge by incorporating the registration of interventional cone-beam computed tomography with standard planning fan-beam CT images in radiotherapy. Accurate image registration is crucial in oncology, particularly for dynamic treatment adjustments in image-guided radiotherapy, where precise alignment is necessary to minimise radiation exposure to healthy tissues while effectively targeting tumours. This work details the methodology and data behind the OncoReg Challenge and provides a comprehensive analysis of the competition entries and results. Findings reveal that feature extraction plays a pivotal role in this registration task. A new method emerging from this challenge demonstrated its versatility, while established approaches continue to perform comparably to newer techniques. Both deep learning and classical approaches still play significant roles in image registration, with the combination of methods, particularly in feature extraction, proving most effective.

URL PDF HTML ☆

赞 0 踩 0