Big Data for Predictive Medicine


Computational Imaging to Predict Intestinal Mucosal Alterations in Children in Virginia

SANA SYED, Asst. Prof. Pediatrics (SOM), DON BROWN, Prof. Systems and Information Engineering (SEAS)

Undernutrition in children can lead to permanent physical and cognitive damage or even death. It affects 20% of children under the age of 5 in low- and middle-income countries, but is also a problem in Virginia, where the percentage of babies with low birth is higher than the national average, even though the percentage of children living in poverty is less than that of the nation. It turns out that the amount of food children eat may not be the only important factor. With funding from the Gates Foundation, scientists at UVA are finding that gastrointestinal infections can also limit the ability of children to process food and absorb nutrients.  

In this project, collaborators from UVA’s Department of Pediatric Gatroenterology and Department of Systems and Information Engineering will develop better methods to identify children who can’t absorb nutrients effectively, so they can be treated. These methods have the potential to improve pediatric nutrition and growth in Virginia and beyond. 

 

 

Delivering Improved Anemia Outcomes in End Stage Renal Disease by Leveraging EMR Data in a Predictive Dosing Algorithm

Brendan Bowman, Asst. Prof, Medicine – Nephrology (SOM), Emaad Abdel-Rahman, Professor, Medicine – Nephrology (SOM), Donald Brown, Professor, Systems & Information Engineering (SEAS)

End Stage Renal Disease (ESRD) is diagnosed when the kidney’s ability to cleanse toxins from blood falls below 15% of normal and can only be treated by kidney transplant or dialysis. While dialysis is a life-saving treatment for ESRD patients, it also has side effects, including a reduction in red blood cell counts called anemia. Anemia is treated using red blood cell stimulating medications (called ESAs), but these drugs are expensive, and individualizing the dose is difficult because it can take several weeks to see the full effects of a change in dose.

The collaborators from the UVA Division of Nephrology and the Department for Systems and Information Engineering, specifically the Data Sciences Institute, aim to provide personalized dosing of ESAs in dialysis patients using a variety of applied data analytics techniques. Using this proposed dosing algorithm could in turn lead to important cost savings in the healthcare system and more consistent red blood cell levels for patients. The UVA dialysis system cares for over 950 patients with ESRD, over 80% of whom receive some form of ESA treatment and would greatly benefit from more personalized medicine.