Supplementary Material for: Use of the Montreal Cognitive Assessment Thai Version to Discriminate Amnestic Mild Cognitive Impairment from Alzheimer’s Disease and Healthy Controls: Machine Learning Results

<b><i>Background:</i></b> The Montreal Cognitive Assessment (MoCA) is an effective and applicable screening instrument to confirm the diagnosis of amnestic mild cognitive impairment (aMCI) from patients with Alzheimer’s disease (AD) and healthy controls (HCs). <b><i>Objectives:</i></b> This study aimed to determine the reliability and validity of the following: (a) Thai translation of the MoCA (MoCA-Thai) and (b) delineate the key features of aMCI based on the MoCA subdomains. <b><i>Methods:</i></b> This study included 60 HCs, 61 aMCI patients, and 60 AD patients. The MoCA-Thai shows adequate psychometric properties including internal consistency, concurrent validity, test-retest validity, and inter-rater reliability. <b><i>Results:</i></b> The MoCA-Thai may be employed as a diagnostic criterion to make the diagnosis of aMCI, whereby aMCI patients are discriminated from HC with an area under the receiver-operating characteristic (AUC-ROC) curve of 0.813 and from AD patients with an AUC-ROC curve of 0.938. The best cutoff scores of the MoCA-Thai to discriminate aMCI from HC is ≤24 and from AD &gt; 16. Neural network analysis showed that (a) aberrations in recall was the most important feature of aMCI versus HC with impairments in language and orientation being the second and third most important features and (b) aberrations in visuospatial skills and executive functions were the most important features of AD versus aMCI and that impairments in recall, language, and orientation but not attention, concentration, and working memory, further discriminated AD from aMCI. <b><i>Conclusions:</i></b> The MoCA-Thai is an appropriate cognitive assessment tool to be used in the Thai population for the diagnosis of aMCI and AD.

**背景：** 蒙特利尔认知评估量表（Montreal Cognitive Assessment, MoCA）是一款高效实用的筛查工具，可用于从阿尔茨海默病（Alzheimer’s disease, AD）患者与健康对照（healthy controls, HCs）人群中甄别出遗忘型轻度认知障碍（amnestic mild cognitive impairment, aMCI）患者，辅助明确其诊断。
**目的：** 本研究旨在验证两项内容的可靠性与有效性：（a）蒙特利尔认知评估量表泰语版（MoCA-Thai）；（b）基于MoCA各子维度界定遗忘型轻度认知障碍的核心特征。
**方法：** 本研究共纳入60名健康对照、61名aMCI患者与60名AD患者。泰语版MoCA具备良好的心理测量学特性，包括内部一致性、同时效度、重测效度与评定者间信度。
**结果：** 泰语版MoCA可作为诊断aMCI的辅助标准，其受试者工作特征曲线下面积（area under the receiver-operating characteristic curve, AUC-ROC）在区分aMCI与健康对照时为0.813，区分aMCI与AD患者时为0.938。用于区分aMCI与健康对照的最佳截断值为≤24分，区分aMCI与AD患者的最佳截断值为>16分。神经网络分析结果显示：（a）相较于健康对照，回忆功能异常是aMCI最核心的特征，语言与定向功能受损分别为第二、第三重要的特征；（b）相较于aMCI患者，AD患者最核心的特征为视空间技能与执行功能异常，而回忆、语言及定向功能受损（而非注意力、集中力与工作记忆损伤）可进一步区分AD与aMCI。
**结论：** 综上，泰语版MoCA是适用于泰国人群的认知评估工具，可用于辅助诊断aMCI与AD。