用注意力图网络融合多维度语言结构，提升阿尔茨海默病检测准确率。
Gated Multi-Graph Fusion via Graph Attention Networks for Alzheimer's Disease Detection
- 构建语义、依存和共现三类图，刻画语言内容-结构-流的多维特征
- 在ADReSSo数据集上达到90.00%准确率，显著优于基线方法
- 通过自适应门控融合机制，有效应对临床症状异质性
自发性言语是阿尔茨海默病（AD）重要的无创生物标志物，但现有系统常忽略语言病理中的非线性结构破坏与临床异质性。本文提出多视图门控图注意力网络，通过自动语音识别（ASR）将音频转录为语义、依存和共现图，基于“内容-结构-流”框架表征言语特征。其中，共现图利用正常语料库的点互信息（PMI）量化叙事逻辑与语言偏离程度。为应对症状多样性，设计自适应门控融合机制动态整合多视图信息。在ADReSSo数据集上的评估显示，模型达到90.00%准确率。消融实验表明，基于PMI的共现图与异质性感知门控对跨人群分类性能至关重要。源代码已公开于https://github.com/opeacc/AD。
原文摘要 · Abstract (English)
Spontaneous speech is a vital non-invasive biomarker for Alzheimer's Disease (AD), yet many systems overlook non-linear structural disruptions and clinical heterogeneity in pathological language. We propose a Multi-View Gated Graph Attention Network that transcribes audio via Automatic Speech Recognition (ASR) to construct semantic, dependency, and co-occurrence graphs, characterizing speech through a "content-structure-flow" framework. Notably, the co-occurrence graph leverages Pointwise Mutual Information (PMI) from a normative corpus to quantify narrative logic and linguistic deviation. To address symptomatic diversity, an adaptive gated fusion mechanism dynamically integrates these views. Evaluated on the ADReSSo dataset, our model achieves 90.00% accuracy. Ablation results confirm that the PMI-based graph and heterogeneity-aware gating are essential for robust classification across diverse clinical populations. Our source code is publicly available at https://github.com/opeacc/AD.
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。