Date on Master's Thesis/Doctoral Dissertation

5-2024

Document Type

Doctoral Dissertation

Degree Name

Ph. D.

Department

Interdisciplinary and Graduate Studies

Degree Program

Interdisciplinary Studies (Individualized Degree), PhD

Committee Chair

Roussel, Jr., Thomas

Committee Member

Behrman, Andrea

Committee Member

Bertocci, Gina

Committee Member

Bertocci, Karen

Committee Member

Quesada, Peter

Author's Keywords

Pediatric spinal cord injury; trunk control; activity-based therapy; rocking chair; muscle activation

Abstract

Introduction: Activity-based locomotor training improves trunk control in children with spinal cord injury (SCI), and there is need for additional activities to allow these children to activate trunk muscles. The purpose of this research was to design, fabricate, and evaluate a rocking chair for children with SCI, and to investigate the activation of muscles during use. Sensorization of the chair to confirm muscle activation is also explored. Methods: Quality Function Deployment (QFD) design methodologies identified and ranked needs and features for the rocking chair. A design was developed and evaluated via finite element analysis. The chair was fabricated and tested for stability and mechanical integrity. Eleven children with SCI and ten typically developing (TD) children aged 2-12 years rocked while surface electromyography was captured. Analyses were performed to compare muscle activity during rocking to baseline and to characterize temporal muscle activation patterns during rocking. Regression analysis and neural networks were used to predict muscle activation during rocking based on data from sensorization of the rocking chair. Results: QFD analysis confirmed the design to have satisfied safety, therapeutic, practical, and aesthetic criteria. Static loading of the prototype chair to 136 kg, and dynamic loading 59 kg confirmed physical integrity, and static and dynamic tip testing showed a tipping factor of safety of 3.6. Analysis of muscle activation during rocking showed a significant increase (p< 0.05) during rocking in both SCI and TD groups confirming a primary hypothesis. Cluster analysis found SCI subgroups (one similar and one dissimilar) compared to TD. Neural networks performed better at predicting muscle activation based on senor data, than regression analysis, with statistically significant correlations (p

Download

Share

COinS