E-Thesis 100 views
Advancing the Real-World Assessment of Movement Quality Using Low-Cost Wearables and Advanced Data Analytics / THOMAS SWAIN
Swansea University Author: THOMAS SWAIN
DOI (Published version): 10.23889/SUThesis.68887
Abstract
Movement quality, an important, yet overlooked component of physical activity, is challenging to assess in real-world settings. This thesis aimed to advance understanding of the assessment and enhancement of movement quality among the general, recreationally active, population using wearable devices...
| Published: |
Swansea University, Wales, UK
2024
|
|---|---|
| Institution: | Swansea University |
| Degree level: | Doctoral |
| Degree name: | Ph.D |
| Supervisor: | Mackintosh, K. A., McNarry, M. A. |
| URI: | https://cronfa.swan.ac.uk/Record/cronfa68887 |
| Abstract: |
Movement quality, an important, yet overlooked component of physical activity, is challenging to assess in real-world settings. This thesis aimed to advance understanding of the assessment and enhancement of movement quality among the general, recreationally active, population using wearable devices and other consumer technologies.Chapter 3 highlighted the improved efficacy of multi-sensor wearables for capturing comprehensive movement data, while showing that multiple devices better accommodate holistic movements. Chapter 3 identified two approaches for analysing movement quality: i) motion-based measurements; and ii) and classification methods. A commercial-grade angular rate, and gravity (MARG) sensor was validated in Chapter 4 for measuring device orientation using sensor-fusion. The sensor’s utility was confirmed under static conditions, though errors increased proportionally with higher angular velocities. Thereafter, Chapter 5 demonstrated the potential of commercial-grade MARG sensors to increase real-world applicability and accessibility of complex biomechanical models, but accentuated the requirement for strategic sensor placement and refined processing methods. Chapter 6 introduced a data augmentation technique that improved class sensitivity and mitigated classification bias in a support vector machine used to determine movement competency. The method addressed the challenges of small, imbalanced datasets, though the implications of subjective data labelling remain. Chapters 7 and 8 collectively highlighted the importance of accessible, user-centric technologies to support movement quality, with wearable-experts (Chapter 7) and the general population (Chapter 8) emphasising the need for convenient technologies and tailored multimodal feedback to address diverse user needs.This thesis demonstrated the potential of commercial-grade wearables to enhance movement quality, dependent on sensor placement, data processing, and feedback. While revealing current limitations in direct motion measurements, the potential for machine learning, supported by data augmentation, was highlighted to facilitate real- world assessments among the recreationally-active population. Multimodal feedback, centred around visualisations, was specifically promoted to maximise movement- quality assessment efficacy and user engagement. |
|---|---|
| Item Description: |
A selection of content is redacted or is partially redacted from this thesis to protect sensitive and personal information. |
| Keywords: |
motor control, motor skill, kinematics, inertial measurement unit, pattern recognition, signal processing, biofeedback, movement proficiency |
| College: |
Faculty of Science and Engineering |
| Funders: |
Polar Electro Oy |