Liang Xiaohui, Batsis John A, Zhu Youxiang, Driesse Tiffany M, Roth Robert M, Kotz David, MacWhinney Brian
Department of Computer Science, University of Massachusetts Boston, 100 Morrissey Blvd., Boston, MA 02125-3393 USA.
Division of Geriatric Medicine, University of North Carolina at Chapel Hill, 5017 Old Clinic Building, Chapel Hill, NC 27599 USA.
Comput Speech Lang. 2022 Mar;72. doi: 10.1016/j.csl.2021.101297. Epub 2021 Sep 22.
Early detection of cognitive decline involved in Alzheimer's Disease and Related Dementias (ADRD) in older adults living alone is essential for developing, planning, and initiating interventions and support systems to improve users' everyday function and quality of life. In this paper, we explore the voice commands using a Voice-Assistant System (VAS), i.e., Amazon Alexa, from 40 older adults who were either Healthy Control (HC) participants or Mild Cognitive Impairment (MCI) participants, age 65 or older. We evaluated the data collected from voice commands, cognitive assessments, and interviews and surveys using a structured protocol. We extracted 163 unique command-relevant features from each participant's use of the VAS. We then built machine-learning models including 1-layer/2-layer neural networks, support vector machines, decision tree, and random forest, for classification and comparison with standard cognitive assessment scores, e.g., Montreal Cognitive Assessment (MoCA). Our classification models using fusion features achieved an accuracy of 68%, and our regression model resulted in a Root-Mean-Square Error (RMSE) score of 3.53. Our Decision Tree (DT) and Random Forest (RF) models using selected features achieved higher classification accuracy 80-90%. Finally, we analyzed the contribution of each feature set to the model output, thus revealing the commands and features most useful in inferring the participants' cognitive status. We found that features of overall performance, features of music-related commands, features of call-related commands, and features from Automatic Speech Recognition (ASR) were the top-four feature sets most impactful on inference accuracy. The results from this controlled study demonstrate the promise of future home-based cognitive assessments using Voice-Assistant Systems.
早期发现独居老年人患阿尔茨海默病及相关痴呆症（ADRD）所涉及的认知衰退，对于制定、规划和启动干预措施及支持系统以改善用户日常功能和生活质量至关重要。在本文中，我们对40名65岁及以上的老年人进行了研究，这些老年人要么是健康对照组（HC）参与者，要么是轻度认知障碍（MCI）参与者，我们使用语音助手系统（VAS），即亚马逊Alexa，来探索语音指令。我们使用结构化协议对从语音指令、认知评估以及访谈和调查中收集的数据进行了评估。我们从每个参与者使用VAS的情况中提取了163个独特的与指令相关的特征。然后，我们构建了包括1层/2层神经网络、支持向量机、决策树和随机森林在内的机器学习模型，用于分类并与标准认知评估分数（如蒙特利尔认知评估（MoCA））进行比较。我们使用融合特征的分类模型达到了68%的准确率，我们的回归模型得出的均方根误差（RMSE）分数为3.53。我们使用选定特征的决策树（DT）和随机森林（RF）模型实现了80 - 90%的更高分类准确率。最后，我们分析了每个特征集对模型输出的贡献，从而揭示了在推断参与者认知状态时最有用的指令和特征。我们发现整体性能特征、与音乐相关的指令特征、与通话相关的指令特征以及自动语音识别（ASR）特征是对推理准确率影响最大的前四个特征集。这项对照研究的结果证明了未来使用语音助手系统进行家庭认知评估的前景。
