Connected speech as an expression of brain health

Spontaneous speech is an expression of who we are. Here, we examine in older adults how connected speech can express brain health and the integrity of cognitive brain circuitry, and indicate subclinical neurodegeneration. We will apply Natural Language Processing (NLP) to characterize speech fragments collected in deeply phenotyped cognitively intact older adults as well as patients with early-stage Alzheimer disease. We will also determine brain mechanisms underlying key NLP processes. Speech recordings will be collected through a secure remote connection in response to questions probing autobiographical memory, episodic memory for personally experienced events and for public events, and semantic knowledge. In the same individuals we will collect volumetric MRI, resting-state functional MRI, as well as amyloid PET imaging. To define brain mechanisms, we will conduct a series of task-related fMRI experiments in young adults.

Keywords:functional neuroanatomy, MRI, Alzheimer biomarkers, neurobiology of language, remote language assessment

Disciplines:Cognitive neuroscience, Natural language processing