Engineers Revolutionize CO2 Conversion: Turning Emissions into Valuable Resources

A new electrode design developed at MIT boosts the efficiency of electrochemical reactions that turn carbon dioxide into ethylene and other products. As the world struggles to reduce greenhouse gas emissions, researchers are seeking practical, economical ways to capture carbon dioxide and convert it into useful products, such as transportation fuels, chemical feedstocks, or even
HomeHealthGroundbreaking AI-Optimized Dataset Unveiled in Type 2 Diabetes Research

Groundbreaking AI-Optimized Dataset Unveiled in Type 2 Diabetes Research

Researchers have announced the release of a significant dataset as part of a groundbreaking study examining biomarkers and environmental factors that may affect the onset of type 2 diabetes. This study involves participants both with and without diabetes, and the preliminary findings suggest a wealth of new information that diverges from earlier research.

On November 8, 2024, researchers have unveiled a key dataset from a significant study regarding biomarkers and environmental influences related to the development of type 2 diabetes. This study includes participants without diabetes as well as those at different stages of the disease, hinting at a rich source of information that goes beyond previous research.

For example, data gathered from unique environmental sensors placed in participants’ homes reveal a notable link between the individuals’ health statuses and their exposure to small air pollutants. Other information collected includes responses from surveys, depression assessments, eye-imaging scans, and conventional measurements of glucose and other biological factors.

All this information is designed for examination by artificial intelligence, aiming to uncover new insights regarding risks, preventive strategies, and the connections between illness and well-being.

“Our data indicates that there is diversity among type 2 diabetes patients; they are not all facing the same challenges. With our comprehensive and detailed datasets, researchers can investigate these variations extensively,” stated Dr. Cecilia Lee, professor of ophthalmology at the University of Washington School of Medicine.

She shared her enthusiasm regarding the high quality of the data collected, which currently includes information from 1,067 individuals, representing just 25% of the targeted total enrollment for the study.

Dr. Lee is the program director of AI-READI (Artificial Intelligence Ready and Equitable Atlas for Diabetes Insights), an initiative supported by the National Institutes of Health that aims to collect and disseminate AI-prepared data for global research efforts in health and disease.

The initial data release is featured in a paper published on November 8 in the journal Nature Metabolism. The authors reaffirm their goal to gather health data from a more racially and ethnically diverse group than prior studies, alongside preparing the data for ethical AI analysis.

“This discovery process has been highly energizing,” said Dr. Aaron Lee, another UW Medicine professor of ophthalmology and the principal investigator for the project. “We are a coalition of seven institutions and multidisciplinary teams that have not collaborated before, but we share the common objective of utilizing unbiased data while ensuring its security as we make it available to researchers worldwide.”

Recruiters at study locations in Seattle, San Diego, and Birmingham, Alabama, are working together to enroll 4,000 participants according to balanced criteria, including:

  • racial/ethnic diversity (1,000 each for white, Black, Hispanic, and Asian participants)
  • disease severity (1,000 each for individuals with no diabetes, prediabetes, those on medication that is non-insulin-controlled, and those on insulin-controlled type 2 diabetes)
  • gender balance (equally male and female)

“Traditionally, scientists have focused on pathogenesis—understanding how individuals become ill—and the related risk factors,” said Aaron Lee. “We aim for our datasets to also be analyzed for salutogenesis, which examines the aspects that contribute to health. For instance, if someone’s diabetes improves, what factors facilitate that recovery? We anticipate that the flagship dataset will reveal new findings about type 2 diabetes from both angles.”

By gathering detailed data from a large group, the researchers hope to construct mimicked health histories illustrating how a person can transition from health to illness and back again.

The data are hosted on a unique online platform and made available in two forms: one set requiring users to agree to specific terms for access and another publicly accessible version that adheres to HIPAA privacy standards.

The pilot data released in summer 2024, involving 204 participants, has already been accessed by over 110 research organizations globally. Researchers are required to authenticate their identity and consent to ethical guidelines before accessing the data. (More information on data access can be found at aireadi.org.)

The AI-READI Consortium includes the University of Washington School of Medicine, University of Alabama at Birmingham, University of California San Diego, California Medical Innovations Institute, Johns Hopkins University, Native Biodata Consortium, Stanford University, and Oregon Health & Science University.

The project is based at the Angie Karalis Johnson Retina Center at UW Medicine, Seattle. Cecilia Lee holds the Klorfine Family Endowed Chair, and Aaron Lee is endowed with the Dan and Irene Hunter Professorship.

This research has been financially supported by the NIH through grants OT2OD032644 and P30 DK035816.