MATRIX AI Consortium at UTSA awarded $200K to study dementia in Hispanic communities

UTSA's Dhireesha Kudithipudi was recently awarded $200,000 from the San Antonio Medical Foundation to apply artificial intelligence models to dementia research serving historically underserved populations

JULY 22, 2021 — Dhireesha Kudithipudi, founding director of the MATRIX AI Consortium for Human Well-Being and the UTSA Robert F. McDermott Endowed Chair in Engineering, was recently awarded $200,000 from the San Antonio Medical Foundation to apply artificial intelligence (AI) models to dementia research serving historically underserved populations, specifically the Hispanic communities in South Texas.

The one-year transdisciplinary project, Disentangling Dementia Patterns Using Artificial Intelligence on Brain Imaging and Electrophysiological Data, focuses on helping the local community through advancements in science.

It is a collaboration between UTSA, UT Health SA, and the South Texas Veterans Health Care System focused on data collection and the development and evaluation of methods for dementia classification using magnetic resonance imaging (MRI) and electroencephalograms (EEG).

“The methods will be applicable across a large number of diseases and populations.”

One of the project’s goals is to create new algorithms to better understand dementia and its subtypes. Additional outcomes include creating a curated collaborative data set that serves as a bridge to development of AI algorithms, exploring if there are cross-correlations with other neurodegenerative diseases, and if new information and correlations can help better classify Alzheimer’s disease (AD).

Alzheimer’s and its related dementias (ADRD) have similarities in terms of clinical presentation, progression, and pattern of neural injury. They are responsible for considerable morbidity, mortality, and health care costs in the United States.

The MATRIX AI team—led by Kudithipudi and supported by graduate research assistant Amirreza Sedaghat and research fellow Zachariah Carmichael—is driving the project while curating the data pipeline and designing and classifying EEG patterns using transfer learning models in AI.

The Biggs Institute for Alzheimer’s and Neurodegenerative Diseases at UT Health San Antonio team—led by Mohamad Habes, along with Felipe S. Salinas, and Gabriel de Erausquin—is collecting and curating the data from new patient populations including those who are underserved, such as the region’s Hispanic community.

The South Texas VA team, led by Adetoun Abisogun Musa, and supported by resident physician Muhammad Samy Abdullah and research nurse Kathleen Franklin, are sharing historical data collected from their extensive patient databases and validating the algorithms in different types of VA populations.

“By curating new AI models, datasets and dementia corpora representing the Hispanic population, the team will also contribute to inclusive AI systems that mitigate biases toward underserved populations and democratize AI in health care,” Kudithipudi explains. “The methods will be applicable across a large number of diseases and populations.”

The study leverages state-of-the-art deep learning methods for early detection of dementia types. The new data will permit the team to develop a new focus on ADRD, focused on designing and validating novel AI algorithms for neurodegenerative disease classification, in addition to understanding patterns of dementia that arise from different pathologies informed by AI algorithms applied to MRI and EEG in Hispanic populations.

ADRD is expected to pose an ever-increasing burden on society due to the gradual aging of the American population. Among ethnicities, Hispanics develop dementia earlier and live with it longer than other population groups in the country. Early detection of dementia, though beneficial for all, is particularly needed in San Antonio where over two-thirds of the population is Hispanic.

UTSA is an urban serving university that aims to tackle society’s grandest challenges. The investment in this project will advance ADRD research as it fills in a much-needed gap with underrepresented data from the Hispanic population.

“The MATRIX AI Consortium, one of UTSA’s newest research institutes, has quite the momentum since launching last year, securing nearly two million in funding in support of impactful research,” said Bernard Arulanandam, vice president for research, economic development, and knowledge enterprise at UTSA. “It’s just the beginning as UTSA propels to become nationally recognized as a world-class research university.”

The MATRIX AI Consortium for Human Well-Being is a transdisciplinary multi-stakeholder research enterprise that serves as a central hub for basic and applied AI research housed at UTSA. MATRIX’s mission is to conduct transformative research in the design, use, and deployment of AI that enhances human life, and to offer rigorous research training opportunities that transcend disciplinary boundaries.

MATRIX consists of 65 researchers spanning across four collaborating organizations: UTSA, UT Health San Antonio, Southwest Research Institute, and Texas Biomedical Research Institute. MATRIX core researchers are recognized thought leaders and have research funding from many prestigious federal and private agencies, including the National Science Foundation, the National Institute of Health, the Air Force Research Laboratory, the U.S. Air Force Office of Scientific Research, the National Aeronautics and Space Administration, and the W.M. Keck Foundation.

— Sarah Hada