SCH: synthetic data and uncertainty quantification for population health - troject summary/abstract the broad, long-term objective of this project is to develop reliable, privacy-preserving, and…
Active Grant Awarded
- Award ID
- RMD021808U1QJJ3
- Recipient
-
North Carolina State University UEI U3NVH931QJJ3
- Awarding agency
-
National Institute for Minority Health and Health Disparities
National Institutes of Health, Department of Health and Human Services
- Obligated
- $313,964
Award facts
- Assistance listing
- 93.1RP NIMHD Research - Research Projects (RP)
- Completion date 0% complete
- Place of performance
- Raleigh, NC 27695, USA
Award details come from USAspending.gov.
About this award
The Department of Health and Human Services National Institutes of Health National Institute for Minority Health and Health Disparities awarded North Carolina State University a $313,964 grant on November 19, 2026, under the NIMHD Research - Research Projects (RP) program (CFDA 93.1RP). The project, "Synthetic Data and Uncertainty Quantification for Population Health," will develop conditional synthetic conformal prediction, a generative-conformal framework that produces synthetic wearable accelerometry trajectories with calibrated uncertainty quantification. Deliverables include conditional generative models for attribute-conditioned accelerometry trajectories, conformal calibration methods that identify unreliable or unsupported synthetic-data-assisted inference, and validation of the framework in WHI/OPACH to estimate behavioral risk factors and their associations with cardiovascular disease, cancer, dementia, mortality, and health disparities. Generative models will be trained and evaluated on wearable data from NHANES and OPACH, with synthetic trajectories serving as auxiliary tools for calibration and uncertainty quantification rather than substitutes for real-world observations. Place of performance is Raleigh, NC 27695, USA. The period of performance runs through November 30, 2030.
Description
SCH: synthetic data and uncertainty quantification for population health - troject summary/abstract the broad, long-term objective of this project is to develop reliable, privacy-preserving, and uncertainty-aware data science methods to support population health decision-making when objective health measurements are limited or unevenly available. Wearable sensors provide rich time-series data on sleep, physical activity, and circadian rhythm, but these data are often available only for small subsets of large cohorts. This limitation restricts research on behavioral risk factors and their associations with cardiovascular disease, cancer, dementia, mortality, and health disparities. This project will develop conditional synthetic conformal prediction, a generative-conformal framework for producing synthetic wearable trajectories with calibrated uncertainty quantification. The specific aims are to: (1) develop conditional generative models for attribute-conditioned accelerometry trajectories; (2) develop conformal calibration methods to quantify uncertainty and identify unreliable or unsupported synthetic-data-assisted inference; and (3) validate the framework in WHI/opach to estimate behavioral risk factors and their associations with major health outcomes. The research design integrates generative modeling, conformal prediction, and population health analytics. Generative models will be trained and evaluated using wearable data from nhanes and opach, with synthetic trajectories used as auxiliary tools for calibration and uncertainty quantification rather than as substitutes for real-world observations. Synthetic outputs will be assessed using distributional diagnostics, biological-plausibility checks, subgroup calibration, and validation against observed opach measures and published WHI/opach findings. By extending calibrated wearable-based inference from monitored subsets to larger cohorts while clearly indicating uncertainty and limits of data support, this project will provide methods, software, and workflows for the responsible use of privacy-preserving synthetic data in population health research. The proposed work is relevant to nimhd's mission because IT will support more rigorous analysis of behavioral risk factors in populations with limited wearable data and inform evidence based strategies to reduce health disparities.
Quick answers
Who received award RMD021808U1QJJ3?
RMD021808U1QJJ3 was awarded to North Carolina State University by National Institute for Minority Health and Health Disparities (National Institutes of Health) on November 19, 2026.
How much is award RMD021808U1QJJ3 worth?
$313,964 has been obligated on RMD021808U1QJJ3.
On GovTribe
See this award's full record
- Every transaction and subaward
- The recipient's other awards and teaming partners
- Similar grant awards
- Ask GovTribe AI about this award