Quick Buzz Feed

Duke initiative aims to bring dementia AI tools into clinical practice

Gary Lloyd | Oct 21,26 | 01:40 EST

Technology

Artificial intelligence (AI) could transform care for older adults and people living with dementia, but few tools have successfully moved from research into routine clinical practice.

Building shared infrastructure Duke University has been awarded nearly $25 million over five years for the Duke Artificial Intelligence and Technology Collaboratory for Aging: Governance and Evaluation (Duke AITC-AIGE). This initiative, one of five national centers funded by the National Institute on Aging, aims to develop robust shared infrastructure, finance research projects across the country, standardize various aging-related databases, and enhance expertise. The ultimate goal is to bridge the gap between promising AI innovations developed in research settings and their successful, routine application in clinical practice for older adults and individuals living with Alzheimer's disease and related dementias. Duke AITC-AIGE is uniquely positioned as the only center led from the Southeast, leveraging Duke's strong capabilities in geriatrics, dementia research, translational science, and responsible AI governance. Making AI real-world ready Dr. Juliessa Pavon emphasizes that the core objective of this work is to ensure AI is 'real-world ready.' This involves creating the necessary infrastructure and fostering collaborations to move AI into practice in a trustworthy manner. While there is no shortage of AI algorithms, there is a significant lack of AI solutions that have undergone rigorous evaluation, safe implementation, continuous monitoring, and demonstrated tangible improvements in real-world outcomes for older adults. The initiative aims to establish a clear pathway for transforming promising AI concepts into reliable and beneficial tools in clinical settings. For people with dementia, trustworthy AI holds potential for early detection of cognitive decline through analysis of speech, behavior, and wearable data, as well as identifying risks like falls, optimizing medication, and providing personalized guidance for caregivers. AI tool transparency The Duke AITC-AIGE, based within the Duke Clinical and Translational Science Institute, will offer critical resources including pilot funding, data science infrastructure, and implementation support, alongside robust responsible AI safeguards. A key innovation will be the development of 'model cards,' which will function similarly to food nutrition labels by providing standardized, clear, and comprehensive information about each AI tool. These cards will detail the tool's intended purpose, the datasets used for its training, its performance metrics during testing, its strengths, and areas where it might be less reliable. Crucially, they will also describe potential biases, inherent limitations, and appropriate use cases. This commitment to transparency aims to empower clinicians, health systems, researchers, and patients to make informed decisions about an AI tool's suitability for specific situations. Furthermore, the Research Innovation Core will advance statistical and AI methods to identify and address performance limitations, while NeuroAI, a secure analytics platform, will facilitate the integration of diverse health data (medical records, brain images, lab results, genetic information) to develop and evaluate AI tools for aging and dementia research. From academia to patients A significant aspect of this initiative is the commitment to 'democratize' AI expertise, making it accessible beyond the confines of major academic research institutions. Through a national pilot funding program, investigators across the United States, including community researchers, smaller institutions, patients, and even caregivers, will gain access to specialized knowledge and support. This includes expertise in AI evaluation, data science, implementation strategies, ethical considerations, governance frameworks, and effective stakeholder engagement. By offering a secure, standardized platform with readily available data and methods, the AITC resources are designed to reduce the typical time, cost, and technical proficiency barriers involved in AI tool development and testing. This inclusive approach encourages broader participation in innovation, ensuring that a wider range of ideas can be explored and brought to fruition, not just those originating from large academic centers. Improving lives The overarching and ultimate goal of the Duke AITC-AIGE initiative is to profoundly improve the lives of older adults and individuals affected by dementia. This will be achieved through the strategic development and deployment of technologies that foster greater independence, provide essential support for caregivers, enhance the efficiency and quality of care delivery, and ultimately elevate the overall quality of life for these populations. The measure of success for this five-year endeavor will be the establishment of a clearer, reproducible, and reliable pathway that effectively translates AI innovation into safe, trustworthy, and genuinely beneficial outcomes for older adults.

Building shared infrastructure

Duke University has been awarded nearly $25 million over five years for the Duke Artificial Intelligence and Technology Collaboratory for Aging: Governance and Evaluation (Duke AITC-AIGE). This initiative, one of five national centers funded by the National Institute on Aging, aims to develop robust shared infrastructure, finance research projects across the country, standardize various aging-related databases, and enhance expertise. The ultimate goal is to bridge the gap between promising AI innovations developed in research settings and their successful, routine application in clinical practice for older adults and individuals living with Alzheimer's disease and related dementias. Duke AITC-AIGE is uniquely positioned as the only center led from the Southeast, leveraging Duke's strong capabilities in geriatrics, dementia research, translational science, and responsible AI governance.

Making AI real-world ready

Dr. Juliessa Pavon emphasizes that the core objective of this work is to ensure AI is 'real-world ready.' This involves creating the necessary infrastructure and fostering collaborations to move AI into practice in a trustworthy manner. While there is no shortage of AI algorithms, there is a significant lack of AI solutions that have undergone rigorous evaluation, safe implementation, continuous monitoring, and demonstrated tangible improvements in real-world outcomes for older adults. The initiative aims to establish a clear pathway for transforming promising AI concepts into reliable and beneficial tools in clinical settings. For people with dementia, trustworthy AI holds potential for early detection of cognitive decline through analysis of speech, behavior, and wearable data, as well as identifying risks like falls, optimizing medication, and providing personalized guidance for caregivers.

AI tool transparency

The Duke AITC-AIGE, based within the Duke Clinical and Translational Science Institute, will offer critical resources including pilot funding, data science infrastructure, and implementation support, alongside robust responsible AI safeguards. A key innovation will be the development of 'model cards,' which will function similarly to food nutrition labels by providing standardized, clear, and comprehensive information about each AI tool. These cards will detail the tool's intended purpose, the datasets used for its training, its performance metrics during testing, its strengths, and areas where it might be less reliable. Crucially, they will also describe potential biases, inherent limitations, and appropriate use cases. This commitment to transparency aims to empower clinicians, health systems, researchers, and patients to make informed decisions about an AI tool's suitability for specific situations. Furthermore, the Research Innovation Core will advance statistical and AI methods to identify and address performance limitations, while NeuroAI, a secure analytics platform, will facilitate the integration of diverse health data (medical records, brain images, lab results, genetic information) to develop and evaluate AI tools for aging and dementia research.

From academia to patients

A significant aspect of this initiative is the commitment to 'democratize' AI expertise, making it accessible beyond the confines of major academic research institutions. Through a national pilot funding program, investigators across the United States, including community researchers, smaller institutions, patients, and even caregivers, will gain access to specialized knowledge and support. This includes expertise in AI evaluation, data science, implementation strategies, ethical considerations, governance frameworks, and effective stakeholder engagement. By offering a secure, standardized platform with readily available data and methods, the AITC resources are designed to reduce the typical time, cost, and technical proficiency barriers involved in AI tool development and testing. This inclusive approach encourages broader participation in innovation, ensuring that a wider range of ideas can be explored and brought to fruition, not just those originating from large academic centers.

Improving lives

The overarching and ultimate goal of the Duke AITC-AIGE initiative is to profoundly improve the lives of older adults and individuals affected by dementia. This will be achieved through the strategic development and deployment of technologies that foster greater independence, provide essential support for caregivers, enhance the efficiency and quality of care delivery, and ultimately elevate the overall quality of life for these populations. The measure of success for this five-year endeavor will be the establishment of a clearer, reproducible, and reliable pathway that effectively translates AI innovation into safe, trustworthy, and genuinely beneficial outcomes for older adults.