Multimodal Ordinal Modeling of Alzheimer's Disease Severity Using Structural MRI and Clinical Data

📄 arXiv: 2606.11794v1 📥 PDF

作者: Boris-Stephan Rauchmann, Jonathan Laib, Buse Ercik, Robert Perneczky, Sergio Altares-López

分类: cs.LG, cs.AI

发布日期: 2026-06-10

备注: 18 pages. Submitted to journal for review

💡 一句话要点

提出多模态序数建模方法以自动评估阿尔茨海默病严重性

🎯 匹配领域: 支柱二：RL算法与架构 (RL & Architecture) 支柱九：具身大模型 (Embodied Foundation Models)

关键词: 阿尔茨海默病 多模态学习 序数回归 机器学习 临床决策支持 神经退行性疾病 可解释性AI

📋 核心要点

  1. 现有的阿尔茨海默病分级方法耗时且易受个体差异影响，缺乏高效的自动化工具。
  2. 本文提出了一种基于注意力机制的多模态机器学习框架，结合MRI影像及临床数据进行序数回归。
  3. 实验结果显示，序数多模态模型在相邻阶段准确率和与临床分级一致性上均有显著提升。

📝 摘要（中文）

神经退行性疾病如阿尔茨海默病（AD）需要准确且可扩展的工具来评估疾病严重性，但现有临床分级方法耗时且易受变异影响。本文提出了一种增强注意力的多模态机器学习框架，结合序数回归实现自动化和可解释的AD严重性分级。该框架整合了T1加权MRI与人口统计和遗传变量，比较了单模态与多模态架构的表现。结果表明，序数模型在捕捉CDR量表的有序结构方面表现更佳，且与临床分级一致性更强。使用Grad CAM++和SHAP进行的可解释性分析支持了模型的透明决策过程。

🔬 方法详解

问题定义：本文旨在解决阿尔茨海默病严重性评估的自动化和可解释性问题，现有方法在时间和一致性上存在不足。

核心思路：通过引入多模态数据（T1加权MRI、人口统计和遗传信息）与序数回归，提升模型的准确性和解释能力。

技术框架：整体架构包括数据预处理、模型训练和验证，采用分层抽样确保数据的独立性，避免数据泄漏。

关键创新：引入序数回归的多模态学习框架，能够更好地捕捉CDR量表的有序结构，与传统的非序数模型相比，提供更一致的临床预测。

关键设计：模型使用了注意力机制，损失函数设计为适应序数回归，网络结构结合了卷积神经网络和全连接层，确保了对多模态数据的有效处理。

📊 实验亮点

实验结果显示，单模态T1加权MRI模型的相邻阶段准确率为0.963，而序数多模态模型的相邻阶段准确率提升至0.970，且与临床分级的一致性显著提高（QWK 0.549），表明该方法在准确性和可解释性上均有显著优势。

🎯 应用场景

该研究为阿尔茨海默病的临床评估提供了一种新的自动化工具，能够在临床实践中实现更高效的疾病严重性分级。未来，该方法有潜力扩展到其他神经退行性疾病的评估中，推动个性化医疗的发展。

📄 摘要（原文）

Neurodegenerative diseases such as Alzheimer's disease (AD) require accurate and scalable tools for assessing disease severity, yet current clinical staging remains time-intensive and prone to variability. We propose an attention-enhanced multimodal machine learning framework with ordinal regression for automated and interpretable AD severity staging. The framework integrates T1-weighted MRI with demographic and genetic variables and compares unimodal and multimodal architectures using ordinal and non-ordinal prediction heads. Models were trained and validated using cohort-stratified splits derived from the ADNI, AIBL, and NIFD datasets. A strictly held-out test set was constructed using subjects excluded from all training, validation, preprocessing, and hyperparameter tuning procedures, with subject-level splitting employed throughout to prevent data leakage. Among unimodal approaches, the T1-weighted MRI model achieved slightly higher adjacent-stage accuracy (0.963) and agreement with clinical staging (QWK 0.444) than the tabular model (QWK 0.433). Integrating imaging, demographic, and genetic information improved overall performance. The multimodal non-ordinal baseline achieved the lowest prediction error (MAE 0.340), whereas the ordinal multimodal model achieved the highest adjacent-stage accuracy (0.970) and strongest agreement with clinical staging (QWK 0.549). These findings indicate that ordinal formulations better capture the ordered structure of the CDR scale and yield predictions more consistent with clinical staging. Explainability analyses using Grad CAM++ and SHAP demonstrated anatomically and clinically plausible model behavior, supporting transparent decision-making. Overall, attention-based multimodal learning with ordinal regression represents a robust, interpretable, and scalable approach for automated AD severity staging and AI-assisted clinical decision support.