Home // eTELEMED 2025, The Seventeenth International Conference on eHealth, Telemedicine, and Social Medicine // View article


Accuracy Evaluation of Computer Vision-based Markerless Human Pose Estimation for Measuring Shoulder Range of Motion

Authors:
Thomas Hellstén
Jonny Karlsson

Keywords: computer vision; markerless; range of motion; shoulder.

Abstract:
Telerehabilitation (TR) requires precise joint Range of Motion (ROM) measurement methods. This study assessed the accuracy of a Computer Vision (CV)-based markerless Human Pose Estimation (HPE) application for active shoulder ROM by comparing it with universal goniometry (UG) in 20 healthy volunteers. The correlation coefficients between the two methods were 0.94 for shoulder extension, 0.83 for adduction, 0.76 for abduction, and 0.67 for flexion, with mean differences ranging from 3.6° in flexion to -7° in adduction. These findings indicate that the markerless CV application is a moderately accurate tool for measuring shoulder ROM.

Pages: 46 to 48

Copyright: Copyright (c) IARIA, 2025

Publication date: May 18, 2025

Published in: conference

ISSN: 2308-4359

ISBN: 978-1-68558-270-8

Location: Nice, France

Dates: from May 18, 2025 to May 22, 2025