Team 25

Machine Learning Software for NASA Biometrics

Team Members:
Kristin Huber
Justin Irby
Karly Espinosa
Sam Skeete

Team Mentors:
Kimia Seyedmadani - Research Operation Integration (ROI) Group at NASA

Keith Tucker - Research Operation Integration (ROI) Group at NASA

Khalil Khoury - Department of Health, AZ

David Hope - DataRobot

YouTube Link:
View the video link below before joining the zoom meeting

Zoom Link:
https://asu.zoom.us/j/83763616448

Abstract

NASA needs a continuous, passive, non-invasive cardiovascular monitoring system for astronauts to better understand and combat the effects of missions to space on the cardiovascular system. NASA astronauts have been found to be at abnormal risk for arterial stiffening and other heart conditions, but research is difficult due to the small sample size. In order to expedite this process and accomplish the most with limited data, StarHeart Biometrics is developing a machine learning driven software to process the data and alert crew members and ground control if a medical emergency arises. The software will integrate with any hardware that can produce measurements of ECG, heart rate, blood pressure, and partial pressure of oxygen. The project is being completed under a quality system approach with multiple design review teams ensuring quality at every step of the design process. Data simulation of the cardiovascular system at standard, stressed, and microgravity conditions is being developed to train a ML algorithm for each scenario. Canvassing and research will lead the project’s UI and API platforms to offer a platform excelling in usability, accuracy, and consistency. The product will be designed to function using Bluetooth and to process raw data in such a way that the output file size will be minimal and available within seconds. Using the constructed simulation to test the program for its processing time, power requirements, and output file size will allow for quick optimization to meet all specifications set forth.