Date of Award

5-1-2022

Degree Type

DPT Project

Degree Name

Doctor of Physical Therapy (DPT)

Advisor

Michelle L. Gutierrez

Abstract

Background: Wearable technology has become a widely utilized method of objective data collection in healthcare. Several barriers currently exist that are limiting their clinical utility. 4-D Smart Wireless Motion Sensors (4-D SWMS) have been developed to offer a less obtrusive wireless sensor that can be used in a variety of settings. The addition of machine learning to captured motion sensor data has the ability to improve classification accuracy in the analysis of gait.

Purpose: The purpose of this pilot study is to determine whether the 4-D SWMS are sensitive enough to accurately differentiate normal gait parameters in healthy individuals.

Methods: Forty-three participants were recruited from The University of Texas at El Paso to participate in the study. Subjects were asked to fill out a health history survey prior to performing several tasks selected from the Functional Gait Assessment (FGA) while wearing 10 4-D SWMS along various landmarks on the body. Motion capture data from each sensor was continuously and wirelessly transmitted for analysis using different neural networks to identify and classify differences in normal gait.

Results: 4-D SWMS used in conjunction with a convolutional neural network (CNN) have the ability to differentiate between normal gait parameters with up to 90% accuracy.

Conclusion: This study demonstrates preliminary data to support the use of 4-D SWMS in conjunction with a CNN to accurately differentiate between gait patterns. The ultimate goal of future research is to detect the subtle abnormalities of patients with mild traumatic brain injury to determine need for earlier intervention.

Download

Share

COinS