Date of Award

2022

Degree Type

Thesis

Degree Name

Master of Science in Electrical Engineering (MSEE)

Department

Electrical, Computer, and Biomedical Engineering

First Advisor

Kunal Mankodiya

Abstract

Growth of “e-health” (electronic health) is unprecedented especially after the COVID-19 pandemic. Wearable technology such as smartwatches and smart finger rings have the capability to remotely monitor and quantify biometric and daily life activity parameters such as heart rate, body temperature, sleep, exercise, sedentary periods, and others from patients when they are in naturalistic environments such as home, school, work, and sports. This wearable technology can also be used to remotely monitor longitudinal symptomatic changes in chronic and acute conditions such as neurological and psychiatric disorders that may require some indicators of psychological stress in daily life settings. Wearable technology is capable of gathering large amounts of biometric data that may be complex to interpret for clinicians. One of the research challenges is how this massive wearable data can be presented to clinicians in a way that it can be easily interpretable and help clinicians make informed decisions. This Master Thesis research aims to (1) develop and deploy system architecture of a web app that visualizes wearable sensor data, and (2) understand the wearable sensor data visualization needs of clinicians. In this study, we present data collected via the Microsoft Smart Band (smartwatch) which includes heart rate and accelerometer. The CarePortal web app was deployed on a local server (Raspberry Pi Model 4B) using Apache 2 configurations. While processing this dataset we encountered challenges relating to (1) Inconsistent date synchronization, (2) Duplicate timestamps, and (3) Participant adherence. The CarePortal web app displayed data for 13 participants in total. Daily and Hourly maximum, minimum, average and standard deviation of Heart Rate and Heart Rate Variability were calculated. We conducted five Interviews, in two phases to understand the needs of clinicians. Data visualizations included several types of graphs such as radar chart, stacked bar plot, scatter plot combined with line plot, simple bar plot, and box plot. Results indicate that overall clinicians preferred aggregate information such as daily heart rate instead of continuous heart rate and want to see trends in the wearable sensor data over a period of time.

Available for download on Friday, May 17, 2024

Share

COinS