发表机构
Georgia Institute of Technology; Emory University; Cairo University（佐治亚理工学院; 埃默里大学; 开罗大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究针对轻度认知障碍老年人心理健康筛查的多模态分析局限，提出TAMI框架，通过时间对齐、缺失感知与可解释性设计，在49名受试者的抑郁、焦虑分类中取得良好效果，且开放式问题可实现与完整访谈相当的筛查性能。
AI 中文摘要
轻度认知障碍（MCI）老年人的抑郁和焦虑常因医疗资源可及性有限而被漏诊。远程临床访谈的多模态分析是一种可扩展的筛查方法，但现有方法存在三方面局限：一是未校正不同分辨率多模态特征间的时间错位，导致产生虚假跨模态关联；二是远程记录存在模态缺失不均的情况，但缺失值常被零填充，使其与有效近零测量值无法区分；三是无法将预测结果联合归因于模态、问题和访谈时刻，限制了细粒度临床解释。我们提出Temporally-Aligned、Missingness-Aware、Interpretable（TAMI）多模态融合框架，该框架将语音、语言、面部及生理特征在问答片段内对齐至共享时间线，编码随时间变化的模态级缺失情况，并基于问题上下文调控融合过程。在对49名MCI老年人的访谈中，TAMI的抑郁分类受试者工作特征曲线下面积（AUROC）为0.68，焦虑分类AUROC为0.69；多模态特征的细粒度时间对齐带来了最大的性能提升（Δ≥0.1）。多级别可解释性分析显示，抑郁分类依赖眼动追踪和开放式问题，焦虑分类依赖眼动追踪和头部姿态，归因在所有问题中均匀分布；仅使用开放式问题的回答（时长5.1分钟）时，抑郁模型AUROC为0.67，与使用完整访谈（时长19分钟）的结果无显著差异（p>0.05）。本研究结果支持围绕开放式问题设计访谈方案，用于MCI老年人的抑郁筛查。
英文摘要
Depression and anxiety in older adults with Mild Cognitive Impairment (MCI) are frequently underdiagnosed due to limited access to care. Multimodal analysis of remote clinical interviews is a scalable screening approach, but existing methods have three limitations. First, they do not correct temporal misalignment across multimodal features extracted at different resolutions, inducing spurious cross-modal associations. Second, remote recordings exhibit uneven modality dropout, but missing values are often zero-filled, making them indistinguishable from valid near-zero measurements. Finally, they do not jointly attribute predictions to modalities, questions, and interview moments, limiting fine-grained clinical interpretation. We propose a Temporally-Aligned, Missingness-Aware, Interpretable (TAMI) multimodal fusion framework. TAMI aligns speech, language, facial, and physiological features within question-answer segments on a shared timeline, encodes modality-level missingness over time, and conditions fusion on question context. In interviews with 49 older adults with MCI, TAMI achieved area under the receiver operating characteristic curve (AUROC) scores of 0.68 (depression) and 0.69 (anxiety). Fine-grained temporal alignment of multimodal features produced the largest performance gain ($Δ{\geq}0.1$). Multi-level interpretability analysis revealed that depression classification relied on eyegaze and open-ended questions, while anxiety classification depended on eyegaze and head pose, with attribution uniformly distributed across questions. Using only responses to the open-ended questions (5.1min), the depression model achieved an AUROC score of 0.67, which was not significantly different from using the full interview (19min) ($p>0.05$). Our findings support designing interview protocols centered on open-ended questions for depression screening in older adults with MCI.
Comments19 pages, submitted to IEEE Transactions on Affective Computing
