UAB aims to boost use of artificial intelligence in biomedical research – University of Alabama at Birmingham

stemadminSeptember 19, 2022Stem Cell ResearchNo comments

The UAB projects will investigate was to improve and increase the use of artificial intelligence in biomedical research.

Jake Chen, Ph.D., Photography: Steve Wood The University of Alabama at Birmingham Informatics Institute has been awarded over $2 million over the next four years as part of the National Institutes of Healths newly launched Bridge2AI program. The Bridge2AI program will accelerate the widespread use of artificial intelligence by the biomedical and behavioral research communities, with $130 million to be invested nationally during this time.

Two UAB-led research teams out of approximately 100 contesting teams nationwide have been selected for funding by this program. One will be led by the UAB Informatics Institute and the other by researchers in the Department of Ophthalmology and Visual Sciences.

UAB teams will participate in two of the four data generation projects, which will create new biomedical and behavioral data sets ready to be used for developing AI technologies. In collaboration with other funded institutions, they will be creating data standards and tools for ensuring data are findable, accessible, interoperable and reusable, or FAIR. Both UAB projects will assist in the development of innovative AI-ready data sets. These data sets will assist in solving various human health challenges, including uncovering how genetic, behavioral and environmental factors can influence a persons physical condition throughout their life with AI.

The UAB Informatics Institute team is led by module principal investigator Jake Chen, Ph.D.; co-principal investigator Ying Ding, Ph.D., of the University of Texas at Austin; operations manager Swathi Thaker, Ph.D., UABs Center for Clinical and Translational Science; and Pamela Payne-Foster, M.D., University of Alabama and CCTS, who will be the head of diversity, equity and inclusion.

CCTS Informatics, in close partnership with the UAB Informatics Institute, will play a fundamental role in guiding the collaborative strategy and implementation of the project across all institutions and multidisciplinary investigators of the CM4AI team. The multi-site effort also leverages the CCTS Partner Network, including at the University of Alabama in Tuscaloosa.

Im very excited for this opportunity because this is the first time UAB is participating in an NIH-sponsored AI-related program, Chen said. The functional genomic data set that we generate will enable biomedical data scientists to develop a multi-scale cell architectural view of stem cells and cancer cells, benefiting basic biomedical research, AI-driven drug discovery and companion diagnostic discoveries.

We view Bridge2AI as a major program paradigm shift by NIH to enlist biomedical data scientists to generate and standardize challenge problem-organized biomedical data at scale, making them ready for broader data science AI researchers outside of biology and medicine to develop new tools and impact future treatments, Chen said.

Although some artificial intelligence is already being used in biomedical research and health care, there have been limitations in its widespread adoption due to challenges of applying AI technologies to diverse data types, for a broader context of use. This has been attributed to various factors, including the insufficient collection of deep multi-modality data sets, inherent data/algorithm biases with inadequate ethical or trustworthy considerations, or lack of proper training to new users. NIHs goal through the Bridge2AI program is to create more accurate analyzations and interpretations of data, while simultaneously reducing perceived biases or inequities.

Our CM4AI team will be building hierarchical cell models for several cell lines, making the data and data analysis available for AI researchers to consume, Chen said. The project will involve 10 universities throughout the United States as well as Canada, with the intention of merging various scientific disciplines with other institutions such as Simon Fraser University in Canada and the workforce development team from Yale.

This process can involve various challenges, from forming the right team to creating a culture so that people can work efficiently together, and thats the incredible importance of teaming, Chen said. Creating a shared mission can provide both a unique and an exciting opportunity. Im hoping, with UABs leadership, the teams will work effectively, complete our milestone goals and create a valuable resource to the AI community.

The teaming module not only ensures our distributed team works effectively through a shared team governance structure, but also ensures that we create a diverse, science-driven culture that spans across different geographical areas, Chen said. The teaming module will develop equity and inclusion guidelines, create knowledge graphs to unify disparate knowledge and information, and use the UAB-created U-BRITE team data science infrastructure to manage diverse aspects of the project.

UABs contributions to the Bridge2AI project will include various levels of expertise that will ensure that the CM4AI data being produced will help bridge the technology adoption gap between developers and the end users.

NIH has issued four challenge team awards for data generation projects, and one award to create a partnered Bridge Center for integration, dissemination and evaluation activities.

Data generation projects will also develop training materials that promote a culture of diversity and the use of ethical practices throughout the data generation process. The Bridge Center will be responsible for integrating activities and knowledge across data generation projects, and disseminating products, best-practices and training materials.