The University at Buffalo Institute for Artificial Intelligence and Data Science (IAD) was established in 2021, a continuation of UB’s longstanding pioneering history of artificial intelligence (AI) research. As AI has changed our world, the IAD has expanded in size and impact to unite UB’s researchers, educators, students and partners.

“Connecting with the IAD ensures that Buffalo Translational Consortium researchers can utilize AI technology to enhance their work and advance health discoveries,” says University at Buffalo Clinical and Translational Science Institute (CTSI) Director Sanjay Sethi, MD, SUNY Distinguished Professor, Senior Associate Dean for Clinical and Translational Research, Jacobs School. “Our CTSI partnership with the institute will continue to grow, as will its impact on our clinical research.”

Siwei Lyu, PhD, SUNY Distinguished Professor and SUNY Empire Innovation Professor, Department of Computer Science and Engineering, School of Engineering and Applied Sciences, is Director of the IAD. An internationally renowned researcher whose interests include digital media forensics, computer vision, and machine learning, Lyu is a CTSI Steering Committee Member, and CTSI Informatics Core Affiliated Faculty.

“Clinical and translational researchers are facing both tremendous opportunities and significant challenges as AI rapidly transforms the research landscape,” explains Lyu. “Our role at UB’s Institute for AI and Data Science is not simply to provide technical expertise, but to serve as a bridge between AI methods and real-world biomedical and health challenges. We help investigators identify where AI can create meaningful impact, avoid common pitfalls, and build collaborations that bring together domain experts and AI researchers.”

Specific services for investigators

Lyu outlines three specific services that have already proven to be successful:

• AI Clinic is a 30-minute evaluation to begin the collaboration process: “The AI Clinic is designed as a low-barrier entry point for researchers who are curious about AI but may not know where to begin. Investigators bring a brief description of their project, and our team provides a rapid assessment of whether AI is appropriate, what level of AI engagement may be needed, and what next steps would be most productive. Sometimes the answer is a sophisticated AI solution, and sometimes the answer is that a simpler statistical or computational approach is more appropriate. Either outcome saves researchers time and resources.”

• Engineering and Analytics Studio is for researchers in need of deliverables: “For projects that require more substantial technical development, the Engineering and Analytics Studio provides hands-on support to help investigators build prototypes, analytical pipelines, demonstrations, and other research deliverables. Our goal is to help move promising ideas from concept to implementation while positioning teams for future external funding.”

• Researchers who know they need AI support are encouraged to bring a new AI problem to the IAD team: “Many of the most exciting advances occur when researchers from different disciplines come together around a new problem. We frequently connect clinical investigators with AI faculty across the university who have complementary expertise. These partnerships often lead to stronger research teams, more competitive proposals, and new scientific directions that neither group would have pursued independently.”

Lyu says researchers do not need to have a fully developed AI project before contacting the IAD team. “In fact, engaging early is often the best approach,” he says. “Some of our most successful engagements begin with a simple consultation. Investigators come to us with a question about AI, and through the process we help them refine the problem, identify collaborators, and develop projects that are substantially stronger than their original concept.”

Future plans

Looking ahead, Lyu believes the IAD will play an increasingly strategic role in helping researchers responsibly adopt AI throughout the research lifecycle: “Beyond consultation and technical support, we will focus on building shared infrastructure, facilitating access to advanced computing resources, supporting large interdisciplinary proposals, and promoting responsible and trustworthy AI practices in biomedical and clinical research.”

Lyu says the IAD has seen significant interest in the use of AI to improve communication between researchers, clinicians, and patients, adding that the institute is currently working with CTSI partners to explore AI-assisted approaches for simplifying clinical and research information into more accessible language while maintaining accuracy and trustworthiness.

When considering the risks and benefits of AI, Lyu encourages researchers to think of it as a tool rather than a replacement for scientific expertise.

“The most successful applications of AI occur when domain experts remain firmly in the driver's seat. Researchers do not need to become AI experts themselves; they need to understand what questions AI can help answer and where its limitations lie. That is exactly where the IAD can help.”

To learn more about the IAD or to start the collaboration process, visit buffalo.edu/ai-data-science.html or write to ub-iad@buffalo.edu.