Home // SENSORCOMM 2018, The Twelfth International Conference on Sensor Technologies and Applications // View article
Early Prediction of Hypoxia Based on Vitals Analysis and Predictive Analytics
Authors:
Vahram Mouradian
Afarin Famili
Alexandra Kozhemiakina
Mahdieh Ashiani
Keywords: hypoxia; SensoSCAN; correlation analysis; predictive modeling
Abstract:
This study investigates the SensoSCAN device for both its health monitoring properties and prediction of different diseases by vital sign analysis. The study’s objective is to develop a probabilistic model for predicting the presence of hypoxia using correlation chains to assess patients’ vital signs gathered by the SensoSCAN. Vital signs, including heart rate, oxygen saturation, activity level, and systolic and diastolic blood pressure, are used to monitor patients’ health conditions. Our functional system helps predict hypoxia in its early stages, when distinctive symptoms are absent and patients may not be aware of the presence of the disease. Analysis is made using dependencies in correlation matrix, constructed correlation chains, and predictive analytics. This study utilizes the hypothesis that hypoxia is an effect of consecutive process of activity where the increase in heart rate and respiration rate correlate with a decrease in oxygen saturation. The ultimate goal is to use mathematical Markov processes and build Markov chains, where elevated heart rate and respiration rate and depressed oxygen saturation are caused by higher activity levels. Alternatively, Markov chains are constituted considering other vitals (either heart rate, respiration rate or oxygen saturation) as independent variables. This system will ultimately assist doctors in assessing patients’ health by defining the main dependencies between human vitals.
Pages: 21 to 24
Copyright: Copyright (c) IARIA, 2018
Publication date: September 16, 2018
Published in: conference
ISSN: 2308-4405
ISBN: 978-1-61208-659-0
Location: Venice, Italy
Dates: from September 16, 2018 to September 20, 2018