Peter Graffy Headshot

Peter Graffy, MPH

PhD Candidate, Biomedical Informatics

American Heart Association Predoctoral Fellow

Luther College | University of Wisconsin Madison | Northwestern University

About Me

Hello! I am a PhD candidate (expected dissertation defense: May 2025) in biomedical informatics at Northwestern University and a predoctoral fellow with the American Heart Association. My work focuses on electronic health record linkage for population health surveillance and dynamic risk prediction using novel ML analytical methods. I've developed computer-aided detection and geolinkage tools that harness protected health information in a HIPAA-compliant manner to improve individual and community health outcomes. My dissertation work evaluates the impact of climate change-related extreme heat events on Chicago through community level assessment and individual level computational phenotypes that integrate EHR and ecological data.


Download My CV

Outside of work, I spend most of my time being active, coaching or playing sports, and with family and friends. I played collegiate soccer and was an internationally-recognized ultimate frisbee player with multiple national championships. I now compete in fitness races such as CrossFit and Hyrox with my sights set on doing a sub-9H Leadville 100 MTB race.


Peter Graffy Playing Ultimate Frisbee Peter Graffy at Hyrox Event

Projects

A tool for geospatial linkage of patient healthcare encounters with Daymet climate variables.

The Daymet DeGAUSS Tool is an advanced resource I developed with fellow PhD candidate Ben Barrett to facilitate the integration of high-resolution weather and climate data with research projects. Using Daymet, a data product from Oak Ridge National Laboratory, this tool allows researchers to quickly access daily surface weather parameters such as temperature, precipitation, and shortwave radiation.

Designed with usability in mind, the tool provides a Docker containerized package that allows users to provide geocoded patient data using HIPAA-compliant DeGAUSS architecture and retrieve Daymet weather data from 1980 to the present at a 1km resolution. It's tailored for epidemiologists, environmental health scientists, and others who need precise weather data for modeling and analysis.

Key Features:

  • Supports batch file uploads for large datasets (5+ million encounters).
  • Automatically retrieves historical weather data for user-defined locations.
  • Efficient integration with R and Python workflows.

Whether you're investigating the effects of climate on public health or conducting environmental assessments, this tool provides robust and accurate data access.

Visit the Daymet DeGAUSS Tool to learn more and explore its capabilities.

Chicago Crime Map Shiny App

If you're a map enthusiast and are interested in Chicago crime (for the near North neighborhoods), look no further. This interactive Shiny app visualizes crime data in Chicago for three northern wards. It allows users to explore crime locations by type, date, and neighborhood using an interactive map. This was originally made for my HOA as a few folks in my building were worried about crime in the area.

The interactive map allows you to see the crime classification, date of the crime, and its description. Data are pulled from the Chicago Data Portal.

Explore the Crime Map

And if you want some further insights into the trends for the area, see another shinyapp I made that looks at trends in certain crime types.

See the crime trends here.

Publications

Here is a sample of some of my latest publications:

Loading publications...

Contact

Email: petergraffy@gmail.com