Estimating Cognitive Function Using Spontaneous Speech in Older People Living in the Community

This thesis presents a novel approach to cognitive function estimation using spontaneous speech, focusing on community-dwelling individuals. By analyzing speech data from older adults and applying advanced natural language processing and machine learning techniques, the study develops a method to differentiate between healthy individuals, those with mild cognitive impairment (MCI), and those with dementia. The research explores the psychological burden of using an AI agent as an evaluator, offering promising insights into its effectiveness as a tool for early detection and intervention in cognitive decline. The findings highlight the potential for AI-driven solutions to enhance cognitive assessment and support clinical decisions, marking a significant step towards integrating AI in healthcare.