Public Health in the Age of AI

April 7, 2025

As environmental exposures like air pollution and wildfire smoke become increasingly common, the effects on human health are often difficult to track and even harder to predict. But in a powerful fusion of technology and public health, Suzanne Tamang, PhD, is harnessing artificial intelligence (AI) and one of the largest medical databases in the country to uncover how these exposures impact autoimmune and rheumatic diseases – and to build better systems of care.

Tamang, an Assistant Professor at Stanford and a computer scientist, partners with the Department of Veterans Affairs (VA) to study a critical population: U.S. veterans. Veterans often experience a unique constellation of exposures and vulnerabilities, making them an ideal group for understanding how environmental stressors intersect with chronic illness. Tamang’s work sits at this intersection, where climate, community, and chronic disease meet.

At the core of her research is a natural language processing (NLP) system capable of scanning over a million clinical notes a day. Patient records often contain rich, detailed information – about social conditions, occupational exposures, and health risks – that isn’t captured in standard medical codes. Her NLP model is able to extract and analyze these unstructured data, surfacing patterns that were previously invisible to researchers.

In collaboration with MDClone, a data analytics company, Tamang has taken this work a step further by generating synthetic patient populations. These are not real patients, but AI-generated representations built from real data. By preserving the statistical patterns of real-world patient information – without compromising any individual’s privacy – these synthetic datasets open the door to powerful new kinds of public health research.

“It was derived from real veterans, although it’s synthetic,” she says. “It’s a new method, based on generative AI, and it allows us to bring in environmental data to explore different ways to link it to population health.”

Synthetic patients allow Tamang’s team to simulate how environmental factors like wildfire smoke or urban air pollution could affect different communities across the U.S., based on geography, health status, or social risk. This approach not only protects patient confidentiality but also allows researchers to test hypotheses and build machine learning prediction models on high performance computing platforms that aren't readily available behind the firewall of the VA or other health systems.

Her research is particularly timely during National Public Health Week, (April 7-13, 2025), which emphasizes community resilience, environmental health, and health equity. Tamang’s work embodies these themes by addressing one of the most urgent public health challenges of our time: the growing burden of climate-sensitive disease.

The Chronic Disease Connection

Chronic autoimmune and rheumatic conditions, such as rheumatoid arthritis, lupus, gout and vasculitis, are notoriously difficult to predict. Disease flares can be triggered by a variety of factors – some genetic, some environmental, and many still unknown. Tamang’s research brings clarity to these questions by integrating environmental exposure data (like particulate matter from wildfire smoke or ozone levels) with clinical patterns observed in real-time patient notes. Her findings have the potential to inform early warning systems for vulnerable populations – especially during periods of poor air quality – and to guide more personalized treatment options.

The synthetic population she’s developing is also demographically and geographically diverse, allowing for a more inclusive understanding of disease risk and helping ensure that tools developed from this research can be applied equitably across communities.

“Environmental exposures don’t affect everyone equally,” she notes. “People in underserved areas often face higher risks, not just because of where they live, but also due to systemic inequities in healthcare access. By incorporating environmental and social data, we can start to design systems that anticipate those disparities and respond more effectively.”

Her population health approach also emphasizes collaboration. Tamang’s team includes environmental scientists, engineers, clinicians, and community stakeholders. Together, they are building an ecosystem of data and tools that health systems can use to better care for patients.

“Our work aims to unite various experts to tackle these significant challenges by providing them with comprehensive datasets,” Tamang says. “This approach marks a critical advancement in our ongoing projects within the new climate and health space.”

Ultimately, her goal is to make public health systems smarter, more adaptive, and more proactive, especially when it comes to protecting those most vulnerable to environmental harm.

Dr. Tamang’s research serves as a reminder of the evolving nature of public health practice. By blending emerging technologies with a deep understanding of environmental and social risk, she is helping us reimagine how we track, treat, and ultimately prevent disease.

*The views and opinions expressed in this article are those of the author(s) and do not necessarily reflect the official policy or position of the U.S. Department of Veterans Affairs or the U.S. government.*