How a $25 Million Federal Investment Plans to Turn Dementia AI Promises Into Clinical Reality

Artificial intelligence tools for dementia care have proliferated in research labs, but few make the leap to clinical practice. Now, Duke University is launching a national infrastructure initiative designed to change that. The Duke Artificial Intelligence and Technology Collaboratory for Aging: Governance and Evaluation (Duke AITC-AIGE) has received nearly $25 million over five years from the National Institute on Aging to build the missing bridge between promising AI research and real-world implementation in aging and dementia care.

The collaboratory is one of five centers selected for the National Institute on Aging's Artificial Intelligence and Technology Collaboratories program and the only one led from the Southeast region. With approximately $4 million annually dedicated to research projects nationwide, the initiative aims to democratize access to AI expertise and provide researchers with pilot funding, data science infrastructure, implementation support, and responsible AI safeguards.

Standardizing the Path from Algorithm to Clinic

A central challenge in clinical AI adoption is the lack of transparent, standardized evaluation methods. Duke AITC-AIGE addresses this by developing "model cards"—similar to nutrition labels—that detail AI tool capabilities, training data, performance metrics, biases, and appropriate use cases. These standardized assessments are designed to enable clinicians to make informed decisions about which AI tools merit adoption.

The initiative is housed within the Duke Clinical and Translational Science Institute and led by three principal investigators: Dr. Juliessa Pavon, an Associate Professor in the Division of Geriatrics overseeing the Administrative Core; Dr. Rishi Kamaleswaran, an Associate Professor of Surgery and Anesthesiology leading the Clinical Implementation and Data Science Core; and Dr. Sheng Luo, a Professor of Biostatistics and Bioinformatics directing the Research Innovation Core.

At least 80% of research funds will support projects focused on Alzheimer's disease and related dementias, with the remainder supporting broader aging research. The collaboratory has developed NeuroAI, a secure analytics and evaluation platform designed to integrate medical records, brain images, laboratory results, and genetic information for comprehensive AI tool testing.

AI Applications on the Horizon

The types of AI tools the collaboratory aims to advance span the continuum of dementia care. Applications in development include early cognitive decline detection through speech and behavior analysis, fall and wandering risk identification systems, medication optimization and deprescribing support, and caregiver monitoring systems.

By providing shared infrastructure and standardized evaluation methods, the initiative creates a clearer pathway for health tech companies to bring dementia AI products to market while potentially reducing development costs and time-to-deployment. For healthcare providers, the transparent evaluation framework offers a way to assess AI tools beyond vendor marketing claims.

Democratizing AI Development

Perhaps most significantly, the collaboratory's model extends beyond major academic medical centers. By democratizing access to AI expertise and infrastructure, the initiative enables community researchers, smaller institutions, and even patients to contribute to AI tool development. This broader participation could lead to more diverse, patient-centered solutions that address real-world needs for maintaining independence and quality of life in dementia care.

As AI continues to generate excitement in healthcare, initiatives like Duke AITC-AIGE represent a critical step toward ensuring that algorithmic promise translates into clinical value—with the safeguards, transparency, and accessibility necessary for trustworthy implementation.

Sources

1. Duke initiative aims to bring dementia AI tools into clinical practice — News-Medical.net