Development of a mobile technology system to measure shoulder range of motion

Student thesis: Master's Thesis

Abstract

In patients with shoulder movement impairment, assessing and monitoring shoulder range of motion is important for determining the severity of impairments due to disease or injury and evaluating the effects of interventions. Current clinical methods of goniometry and visual estimation require an experienced user and suffer from low inter-rater reliability. More sophisticated techniques such as optical or electromagnetic motion capture exist but are expensive and restricted to a specialised laboratory environment.Inertial measurement units (IMU), such as those within smartphones and smartwatches, show promise as tools bridge the gap between laboratory and clinical techniques and accurately measure shoulder range of motion during both clinic assessments and in daily life.This study aims to develop an Android mobile application for both a smartphone and a smartwatch to assess shoulder range of motion. Initial performance characterisation of the inertial sensing capabilities of both a smartwatch and smartphone running the application was conducted against an industrial inclinometer, free-swinging pendulum and custom-built servo-powered gimbal.An initial validation study comparing the smartwatch application with a universal goniometer for shoulder ROM assessment was conducted with twenty healthy participants. An impaired condition was simulated by applying kinesiology tape across the participants shoulder girdle. Agreement, intra and inter-day reliability were assessed in both the healthy and impaired states.Both the phone and watch performed with acceptable accuracy and repeatability during static (within ±1.1°) and dynamic conditions where it was strongly correlated to the pendulum and gimbal data (ICC > 0.9). Both devices could perform accurately within optimal responsiveness range of angular velocities compliant with humerus movement during activities of daily living (frequency response of 377°/s and 358°/s for the phone and watch respectively).The concurrent agreement between the watch and the goniometer was high in both healthy and impaired states (ICC > 0.8) and between measurement days (ICC > 0.8). The mean absolute difference between the watch and the goniometer were within the accepted minimal clinically important difference for shoulder movement (5.11° to 10.58°).The results show promise for the use of the developed Android application to be used as a goniometry tool for assessment of shoulder ROM. However, the limits of agreement across all the tests fell out with the acceptable margin and further investigation is required to determine validity. Evaluation of validity in clinical impairment patients is also required to assess the feasibility of the use of the application in clinical practice.
Date of Award1 Nov 2018
LanguageEnglish
Awarding Institution
  • University Of Strathclyde
SponsorsUniversity of Strathclyde
SupervisorMario Giardini (Supervisor) & Philip Rowe (Supervisor)

Cite this