Skip to main content

Seed Grants Will Support U Data Science Research

Feb 10, 2023
Scientist examining data on a computer
The One Utah Data Science Hub pilot seed grant program is part of a university-wide effort to enhance research, training, and infrastructure focused on data science. In 2023, seven University projects will be supported by these grants. Photo credit: Charlie Ehlert/University of Utah Health

Seven University of Utah projects have received seed grants designed to enhance research and infrastructure in data science and data-enabled science. The grants, supported by the new One Utah Data Science Hub, will focus on projects utilizing methods including machine learning, artificial intelligence, and visualization toward solving societally relevant problems within basic and health sciences.

The One Utah Data Science Hub pilot seed grant program is part of a university-wide effort to enhance research, training, and infrastructure focused on datascience. The Hub facilitates cross-campus and interdisciplinary research that focuses on data science through the launch of two new initiatives in alignment with the Utah Center for Data Science:

  1. The Data Science and Ethics of Technology (DATASET) Initiative, and
  2. The Data Exploration and Learning for Precision Health Intelligence (DELPHI) Initiative.

Data science is an umbrella term that encompasses data management, data analytics, data mining, machine learning, visualization, and several other related disciplines. It relies on a multidisciplinary approach to detect and analyze patterns in large amounts of information. It can be applied to detect patterns of disease and improve patient outcomes; better distribute critical health care, food, and supplies during emergencies; or predict energy demands to achieve greater efficiencies and reduce environmental impacts.

"We are thrilled about these diverse data science research projects because of their clear potential for innovation and the exciting new collaborative research teams that have been formed," says Aaron Quinlan, Ph.D., co-director of the DELPHI Initiative and professor of Human Genetics.

The DELPHI Initiative aims to drive innovation in health and medicine by expanding data science expertise and accelerating scientific discovery. The DATASET Initiative aims to bring together research and expertise in all dimensions of data across the U to critically examine the function and impact of data, data infrastructure, data science in addressing grand challenges, science and engineering, and contributing to data-driven decision-making that impacts society. 

Seed grant projects will receive up to $50,000 for one year. 

Project Titles, Summaries, & Awardees

Individualization of Fetal Growth Assessment using Maternal Genetics and Explainable AI

Nathan Blue  M.D. (obstetrics and gynecology)

Mark Yandell Ph.D. (human genetics) Martin Tristani-Firouzi M.D. (pediatrics)

  • Using artificial intelligence, Blue and his colleagues seek to develop a new neonatal morbidity risk calculator that will track fetal growth and pave the way to better-informed decision-making by families and clinicians in complex, high-risk obstetric scenarios.

A Novel Approach to Visualizing Pollution Exposure Patterns in Pregnant Women

Simon Brewer  Ph.D. (geography)

Michelle Debbink  M.D, Ph.D. (obstetrics and gynecology) Brenna Kelly (geography)

  • The researchers will use a machine learning technique to generate self-assembling maps that will help health care professionals visualize the interactions of various air pollutants and their potential influence on pregnancy.

Optimizing Across the Rashomon Set

Nina de Lacy M.D., MBA (psychiatry)

  • This study will delve into the Rashomon Effect, which occurs in machine learning when various models produce the same result but with significantly different explanations. The researchers will develop innovative artificial intelligence approaches to understand the Rashomon Effect better and reduce its influence on research findings.

Predicting Perturbation Phenotypes in the Vertebrate Brain

James Gagnon  Ph.D. (biology)

Randall Peterson Ph.D. (pharmacology and toxicology)

  • Based on studies of zebrafish, Gagnon and colleagues will use machine learning to process and interpret data to better understand the relationships between brain building and brain function.

Use of Modularity Optimization to Define and Evaluate Regional Networks for Emergency General Surgery Care

Marta McCrum  M.D., MPH (surgery)

Joshua J. Horns  Ph.D. (surgery)

Neng Wan  Ph.D. (geography)

  • Emergency General Surgery (EGS) accounts for more than three million hospital admissions each year and over half of all surgical mortality. Using an innovative network science method, McCrum plans to identify and visualize patterns of EGS care in four large states using datasets for emergency and inpatient care. The results could lead to better EGS care and improve patient outcomes.

CURATE Sepsis: CURating A DaTabase from the Electronic Health Records of Patients At-Risk for Sepsis

Daniel Scharfstein ScD (population health sciences)

Ithan Peltan  M.D. MS (internal medicine)

Daniel Knox M.D. (internal medicine)

  • Scharfstein will lead an interdisciplinary team of experts to assemble a unique dataset that contains the electronic health records (EHR) of about 100,000 adults with suspected sepsis admitted to the emergency departments at Intermountain hospitals over a four-year period. Analysis of the data set could help improve treatment of patients suspected of having sepsis.

Machine Learning-Based Heterogeneous Treatment Effects Estimation of 2nd-line Medication Options for Type 2 Diabetes Patients Using Veteran Affairs Electronic Health Record Data

Jincheng Shen  Ph.D. (population health sciences)

Srinivasan Beddhu M.D. (internal medicine) Tom Greene Ph.D. (population health sciences)

  • Utilizing the Veteran Affairs (VA) Electronic Health Record (EHR) system, Shen and colleagues seek to develop new machine learning-based approaches to predict which treatments might be used to reduce the risk of hypoglycemia among individuals with type 2 diabetes. They also hope to identify which patients derive the greatest benefit from each treatment strategy.