Georgia Tech is one of four universities collaborating on a NSF-sponsored project to develop methods and algorithms to turn enormous clinical health record databases into useful phenotypes
Ever since the adoption of electronic health records (EHRs), medical universities, hospitals and other health institutions have amassed enormous databases of information, comprising a diverse array of information such as diagnoses, medications and lab results. While such databases promise to serve as rich resources for clinical research, the data tends to be difficult, time-extensive and costly to analyze. A new project funded by the National Science Foundation (NSF) aims to change that.
As available now, databases of electronic health records are diverse and massive, but they are also messy and heterogeneous. There’s a lot of noise,” said Jimeng Sun, associate professor at Georgia Tech’s School of Computational Science and Engineering. “Our charge is to find ways to make the information more robust and easier to read, thus leading to meaningful clinical concepts without extensive labor and time.”
As part of the four-year, $2.1 million NSF research project, data analytic teams from Georgia Tech and the University of Texas, Austin, will develop algorithms and methods to convert the EHR data into meaningful clinical concepts or phenotypes focused on diseases and specific health traits. Vanderbilt University will provide initial EHR data and phenotype validation. Northwestern University will provide additional data to refine and adapt those phenotypes.
Visit Georgia Tech booth#2857 at SC14.
Sign up for our insideHPC Newsletter.
