Accurate Prediction of Knee Angles during Open-Chain Rehabilitation Exercises Using a Wearable Array of Nanocomposite Stretch Sensors

David S. Wood; Kurt Jensen; Allison Crane; Hyunwook Lee; Hayden Dennis; Joshua Gladwell; Anne Shurtz; David T. Fullwood; Matthew K. Seeley; Ulrike H. Mitchell; William F. Christensen; Anton E. Bowden

doi:10.3390/s22072499

Sensors (Mar 2022)

Accurate Prediction of Knee Angles during Open-Chain Rehabilitation Exercises Using a Wearable Array of Nanocomposite Stretch Sensors

David S. Wood,
Kurt Jensen,
Allison Crane,
Hyunwook Lee,
Hayden Dennis,
Joshua Gladwell,
Anne Shurtz,
David T. Fullwood,
Matthew K. Seeley,
Ulrike H. Mitchell,
William F. Christensen,
Anton E. Bowden

Affiliations

David S. Wood: Department of Mechanical Engineering, Brigham Young University, Provo, UT 84602, USA
Kurt Jensen: Department of Mechanical Engineering, Brigham Young University, Provo, UT 84602, USA
Allison Crane: Department of Mechanical Engineering, Brigham Young University, Provo, UT 84602, USA
Hyunwook Lee: Department of Exercise Science, Brigham Young University, Provo, UT 84602, USA
Hayden Dennis: Department of Exercise Science, Brigham Young University, Provo, UT 84602, USA
Joshua Gladwell: Department of Statistics, Brigham Young University, Provo, UT 84602, USA
Anne Shurtz: Department of Statistics, Brigham Young University, Provo, UT 84602, USA
David T. Fullwood: Department of Mechanical Engineering, Brigham Young University, Provo, UT 84602, USA
Matthew K. Seeley: Department of Exercise Science, Brigham Young University, Provo, UT 84602, USA
Ulrike H. Mitchell: Department of Exercise Science, Brigham Young University, Provo, UT 84602, USA
William F. Christensen: Department of Statistics, Brigham Young University, Provo, UT 84602, USA
Anton E. Bowden: Department of Mechanical Engineering, Brigham Young University, Provo, UT 84602, USA

DOI: https://doi.org/10.3390/s22072499
Journal volume & issue: Vol. 22, no. 7
p. 2499

Abstract

Read online

In this work, a knee sleeve is presented for application in physical therapy applications relating to knee rehabilitation. The device is instrumented with sixteen piezoresistive sensors to measure knee angles during exercise, and can support at-home rehabilitation methods. The development of the device is presented. Testing was performed on eighteen subjects, and knee angles were predicted using a machine learning regressor. Subject-specific and device-specific models are analyzed and presented. Subject-specific models average root mean square errors of 7.6 and 1.8 degrees for flexion/extension and internal/external rotation, respectively. Device-specific models average root mean square errors of 12.6 and 3.5 degrees for flexion/extension and internal/external rotation, respectively. The device presented in this work proved to be a repeatable, reusable, low-cost device that can adequately model the knee’s flexion/extension and internal/external rotation angles for rehabilitation purposes.

Published in Sensors

ISSN: 1424-8220 (Online)
Publisher: MDPI AG
Country of publisher: Switzerland
LCC subjects: Technology: Chemical technology
Website: http://www.mdpi.com/journal/sensors

About the journal

Abstract

Keywords