Screening for Alzheimer's Disease in Mainland China Based on Syntax Impairment

The present study aimed to determine whether linguistic impairments could be used to discriminate individuals with Alzheimer's disease (AD) from healthy controls (HC). Native Chinese speakers with AD, mild cognitive impairment (MCI) and healthy controls were asked to describe the famous Cookie Theft picture. The spontaneous speech was transcribed into text, and Natural Language Processing (NLP) techniques were employed to extract linguistic features. The findings showed significant impairments in both MCI and AD groups compared to the HC group across several word and sentence-level features. Notably, the study demonstrated a significant ability to differentiate between HC, MCI and AD groups, achieving an accuracy of 95% between HC and AD, and 82% between HC and MCI. Syntactic features outperformed lexical features, indicating that linguistic impairment is a valid tool that may be useful for AD screening.