Inertial measurement units (IMUs) are being recognized as a portable and cost-effective alternative to motion analysis systems and have the potential to be introduced into clinical settings for the assessment of functional movement quality of the spine in patients with low back pain. However, uncertainties regarding sensor accuracy and reliability are limiting the widespread use and acceptance of IMU-based assessments into routine clinical practice. The objective of this work was to assess the performance of inexpensive wearable IMUs (Mbientlab MetaMotionR IMUs; Mbientlab Inc., San Francisco, USA; product specifications available in Appendix C) relative to conventional optical motion capture equipment (Vicon Motion Systems Ltd., Oxford, UK) in: 1) a controlled environment, and 2) an uncontrolled environment.
The first study evaluated the performance of 2 IMUs in a controlled environment during simulated repetitive spine motion carried out by means of a motorized gimbal. Root mean square error (RMSE) and mean absolute measurement differences between cycle-to-cycle minimum, maximum, and range of motion values, as well as correlational analyses within IMUs and between IMUs and Vicon, in all movement directions (i.e., simulated flexion-extension (FE), lateral bending (LB), and axial twisting (AT)), were compared. Measurement error was low in all axes during all tests (i.e., ≤ 1.54°); however, low-to-moderate correlational results were found in one non-primary axis, and this axis changed depending on the direction of the movement (i.e., LB during FE-motion (0.244 ≤ R ≤ 0.515), AT during LB-motion (0.594 ≤ R ≤ 0.795), and FE during AT-motion (0.002 ≤ R ≤ 0.255)).
The second study was designed to assess the performance of the IMUs in an uncontrolled environment during repetitive spine FE in human participants. Absolute angles and local dynamic stability were compared for individual IMUs (which were placed over T10-T12 spinous processes, and the pelvis) as well as for relative motion between IMUs. Maximum finite-time Lyapunov exponents (λmax) were used to quantify local dynamic stability and were calculated using both FE and the sum of squares (SS) from measured spine kinematics. It was found that the IMUs have acceptable performance in all axes when tracking motion (RMSE ≤ 2.43°); however, low-to-moderate correlational results were found in one non-primary axis (0.987 ≤ RFE ≤ 0.998; 0.746 ≤ RLB ≤ 0.978; 0.343 ≤ RAT ≤ 0.679). In addition, correlations between λmax estimates were high; therefore, local dynamic stability can be accurately estimated using both FE and SS data (0.807 ≤ 〖ICC〗_2,1^FE ≤ 0.919; 0.738 ≤ 〖ICC〗_2,1^SS ≤ 0.868). Correlation between λmax estimates was higher when using FE data for individual sensors/rigid-body marker clusters; however, correlation was higher when using SS data for relative motion.
In general, the results of these studies show that the MetaMotionR IMUs have acceptable performance in all axes when considering absolute angle orientation and motion tracking, and measurement of local dynamic stability; however, there is low-to-moderate correlation in one non-primary axis, and that axis changes depending on the direction of motion. Future research will investigate how to optimize performance of the third axis for motion tracking; it will also focus on understanding the significance of the third axis performance when calculating specific outcome measures related to spine movement quality.
Identifer | oai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/38698 |
Date | 15 January 2019 |
Creators | Beange, Kristen |
Contributors | Graham, Ryan, Chan, Adrian |
Publisher | Université d'Ottawa / University of Ottawa |
Source Sets | Université d’Ottawa |
Language | English |
Detected Language | English |
Type | Thesis |
Format | application/pdf |
Page generated in 0.0016 seconds