News
Article
AI-ready dataset released in type 2 diabetes study
Author(s):
Key Takeaways
- The study provides a comprehensive dataset on type 2 diabetes, including biomarkers and environmental influences, analyzed using AI for novel insights.
- Data reveals a link between air pollution and diabetes progression, highlighting environmental factors' role in disease development.
Early results suggest participant diversity and novel measures will enable new, artificial intelligence-driven insights. As the study evolves, it could lead to significant advancements in understanding how environmental exposures contribute not only to diabetes but also to its complications, including those affecting eye health, such as diabetic retinopathy.
(Image credit: Adobe Stock/Orawan)
A new study is releasing its flagship dataset, offering a wealth of data on the biomarkers and environmental influences that may play a role in the development of type 2 diabetes. By including participants ranging from healthy individuals to those at various stages of the disease, the research provides a unique, multifaceted perspective that goes beyond traditional diabetes studies.
One of the early findings comes from data gathered through customized environmental sensors placed in participants' homes. The results reveal a clear link between disease progression and exposure to fine particulate matter, a form of air pollution known to have detrimental effects on health. In addition to these environmental factors, the dataset includes a variety of other measures, such as eye-imaging scans, depression scales, traditional glucose measurements, and survey responses, creating a comprehensive picture of each participant's health.1
The data will be analyzed using artificial intelligence to uncover novel insights into the risk factors, preventive strategies, and underlying mechanisms that connect disease and health. As the study evolves, it could lead to significant advancements in understanding how environmental exposures contribute not only to diabetes but also to its complications, including those affecting eye health, such as diabetic retinopathy.
“We see data supporting heterogeneity among type 2 diabetes patients — that people aren’t all dealing with the same thing. And because we’re getting such large, granular datasets, researchers will be able to explore this deeply,” said Cecilia S. Lee, MD, professor of ophthalmology at the University of Washington School of Medicine.
Lee expressed excitement at the quality of the collected data, which represent 1,067 people, just 25% of the study’s total expected enrollees.
Lee is program director of (AI-READY) Artificial Intelligence Ready and Equitable Atlas for Diabetes Insights).2 The National Institutes of Health-supported initiative aims to collect and share AI-ready data for scientists worldwide to analyze for new clues about health and disease.
The initial data release is highlighted in a paper published in the journal Nature Metabolism.3 The authors restated their aim to gather health information from a more racially and ethnically diverse population than has been measured previously, and to make the resulting data ready, technically and ethically, for AI mining.
Aaron Lee, MD, also a UW Medicine professor of ophthalmology and the project’s principal investigator, said the process of discovery has been invigorating.
“We’re a consortium of seven institutions and multidisciplinary teams that had not worked together before,” he said. “But we have shared goals of drawing on unbiased data and protecting the security of that data as we make it accessible to colleagues everywhere.”
At study sites in Seattle, San Diego, and Birmingham, Alabama, recruiters are collectively enrolling 4,000 participants, with inclusion criteria promoting balance:
- race/ethnicity (1000 each – white, Black, Hispanic and Asian)
- disease severity (1000 each – no diabetes, prediabetes, medication/non-insulin-controlled and insulin-controlled type 2 diabetes)
- sex (equal male/female split)
Moreover, Aaron Lee pointed out that scientists are examining pathogenesis — how people become diseased — and risk factors.
“We want our datasets to also be studied for salutogenesis, or factors that contribute to health,” he said. “So, if your diabetes gets better, what factors might be contributing to that? We expect that the flagship dataset will lead to novel discoveries about type 2 diabetes in both of these ways.”
In conclusion, the goal of this ambitious study is to create detailed "pseudo health histories" that chart the progression from disease to health and vice versa, providing invaluable insights into the complex factors that influence type 2 diabetes and its complications. By gathering richly detailed data from a large cohort of participants, the researchers aim to build a more nuanced understanding of how both environmental and biological factors interact over time.
To facilitate collaboration and further research, the dataset is hosted on a custom online platform and is available in two forms: a controlled-access set, which requires a usage agreement, and a publicly accessible version that strips away HIPAA-protected information. The pilot release in summer 2024, involving data from 204 participants, has already been downloaded by over 110 research organizations worldwide. To access the data, researchers must verify their identity and agree to ethical usage terms. For more information on how to access the dataset, visit aireadi.org.
This collaborative offers insights into the prevention and management of type 2 diabetes, with potential applications in the prevention of diabetic eye diseases.1
