NIH Awards Zolnoori $3M for AI Study Detecting Cognitive Decline Early in Black Older Adults
Maryam Zolnoori, PhD, an assistant professor at Columbia Nursing, has received $3,406,699 from the National Institute on Aging of the National Institutes of Health (NIH) to develop an approach to detect cognitive decline in Black older adults early using artificial intelligence (AI).
The project, titled “Tackling Disparity with Sound: Audio-Recorded Patient-Clinician Communication for Early Detection of Mild Cognitive Impairment in Black Older Adults,” will create a large database of conversations between Black patients and nurses to train AI models capable of identifying early signs of cognitive impairment.
Mild cognitive impairment and early-stage dementia affect approximately one in five adults over the age of 60. Over 50% of these cases remain undiagnosed—63% within home health care (HHC). Black older adults are at higher risk of delayed or missed diagnosis due to structural barriers in health care access and biases in traditional cognitive assessments.
Subtle changes in language and speech are one of the earliest signs of cognitive decline. Advances in natural language processing, an AI branch that analyzes human language, make it possible to detect these changes by examining patient–clinician conversations alongside information from electronic health records. However, current AI models are trained on speech datasets that do not adequately represent Black verbal communication patterns which can lead to misinterpretation and delayed care.
In collaboration with VNS Health, one of the nation’s largest home- and community-based health care nonprofits, Zolnoori and her team will compile the largest dataset of Black patient-nurse conversations in home health care. The data will be used to develop and refine AI models capable of accurately analyzing linguistic patterns and detecting early signals of cognitive decline.
The project will also evaluate how these models can be integrated into home health care workflows to support earlier identification of cognitive impairment and facilitate timely referral for clinical evaluation and care planning.