A Multimodal Fusion Framework for Early Detection of Cognitive Impairment in Chinese Speakers Using Pinyin Sequences and Acoustic Features

Abstract

Background: Alzheimer's disease (AD) is a leading cause of dementia, and traditional diagnostic methods like cerebrospinal fluid testing and PET imaging are invasive, costly, and limit early detection. Language biomarker analysis offers a non‐invasive, efficient alternative to detect cognitive impairments through speech.

The paper