Core Body Temperature Estimation to Detect Ebola Virus Disease

Name: Bill Smart
Affiliation: Oregon State University
Phone: (541) 737-0670
Knowledge Required: Mathematical modeling, statistics, machine learning.
Motivation: This project is part of a larger project looking at using robotics and automation in the fight against Ebola Virus Disease. The project is funded by the National Institutes of Health, and we are collaborating with Medecins Sans Frontieres (Doctors Without Borders, MSF) in Brussels, Belgium. MSF led the response to the 2015-17 Ebola outbreak in West Africa.

Having an automated triage system that works reliably in the field will allow them to dramatically reduce the risk of infection for both their own staff and for members of the public coming in for treatment. If this project is successful and ends up being deployed in the field, there is a real chance that it will save lives.
Description: We are working with Medecins Sans Frontieres (Doctors without Borders, MSF) to investigate how automation can help in the fight against Ebola Virus Disease. One thing that would really help is to be able to quickly triage patients coming into a hospital facility, identifying those with elevated core body temperatures. The goal of this project is to build a mathematical model that predicts core body temperature, based on skin temperature and other sensor readings.
Objectives: The goal of this project is to build a computational model that predicts the core body temperature of individuals using only stand-off sensors, like a thermal imaging camera. This system will be used to triage patients arriving at Ebola treatment centers, allowing those who are not symptomatic (normal body temperature) into the center, while diverting those who may be symptomatic (high body temperature) into the treatment areas. The physical sensor system will be designed and fabricated as part of an MIME senior design project, and you will have some design input into what kinds of sensors you need, in order to improve model quality.

People who are showing signs of Ebola Virus Disease have an elevated core body temperature. This is fairly easy to detect with contact sensors (thermometers, for example), but this places care workers at risk of infection, and does not scale to large numbers of people (since it takes minutes to take the readings). The other complication is that the skin temperature, which can be estimated by a thermal camera, is not the same as the core temperature. It is affected by ambient temperature, sweating, whether you just ran up the stairs, and a number of other confounding factors.
Deliverables: A computational model that accurately predicts core body temperature, based on data from stand-off sensors. We will collect a set of ground-truth data (stand-off sensor readings and actual core body temperatures), and use this to verify that the system works as it should, and to establish uncertainty bounds on the estimates that it produces.
Other comments:

   D. Kevin McGrath
   Last modified: Tue Mar 6 10:29:08 2018