Assessing movement quality on straight leg raise using neural networks and data science

Dopazo, D.A. Adanza; Button, K.B.; Gardner, S.G.; Al-Amri, M.A.

Abstract

Wearable sensors used to measure position/orientation and acceleration during exercise for knee and hip osteoarthritis have the potential to enhance physiotherapy rehabilitation through the personalisation of exercise. This data can be used to monitor exercise performance from the home and provide personalised feedback based on the quality of movement during the exercises outside of the clinical setting. Data science can be implemented to objectively characterise the quality of the movement patterns. This is achieved with intelligent algorithms that identify the movement quality based on orientations and acceleration data, which represent common feedback given by physiotherapists. These algorithms aim to recognize the underlying relationships inside the data, emulating the way the human brain functions. This results in a classification system that can distinguish between a good and a difficult movement.This study aimed to develop a neural network to assess the quality of movement during one rehabilitation exercise, the straight leg raise. This exercise was selected because it is a commonly prescribed non-weight bearing exercise used early in a rehabilitation programme.

• Journal Article Type: Article
• Acceptance Date: Feb 1, 2022
• Online Publication Date: Mar 28, 2022
• Publication Date: Mar 28, 2022
• Deposit Date: Aug 22, 2022
• Publicly Available Date: Mar 29, 2023
• Journal: Osteoarthritis and Cartilage
• Print ISSN: 1063-4584
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 30
• Issue: Supplement 1
• Pages: S93
• DOI: https://doi.org/10.1016/j.joca.2022.02.116
• Keywords: Orthopedics and Sports Medicine; Biomedical Engineering; Rheumatology
• Public URL: https://uwe-repository.worktribe.com/output/9852467
• Publisher URL: https://www.sciencedirect.com/science/article/pii/S1063458422001509?via%3Dihub