Prediction of Cognitive Test Performance Using AI-Based Analysis of Narrative Speech

About this study

This study aims to evaluate a new artificial intelligence (AI)-based method for measuring cognitive function using speech recordings. Participants will complete a short storytelling task in which they describe a story based on an image while their voice is recorded using a computer or mobile device. The speech recordings will be analyzed using AI technology to identify patterns in speech that may be related to cognitive function. The system will then estimate scores that correspond to commonly used cognitive tests. To evaluate the accuracy of this method, the AI-generated scores will be compared with results from standard cognitive assessments administered by trained researchers. These assessments may include tests commonly used to measure memory, attention, and other cognitive abilities. The goal of this study is to determine whether speech analysis using AI can provide a convenient and efficient approach for cognitive assessment. If successful, this technology may help support early detection of cognitive decline and provide a practical tool for large-scale or remote cognitive screening.

Sponsor: Chang Gung Memorial Hospital