Accuracy and repeatability of single-pose calibration of inertial measurement units for whole-body motion analysis

摘要

Portable inertial measurement units (IMUs) are suitable for motion analysis outside the laboratory. However, IMUs depend on the calibration of each body segment to measure human movement. Different calibration approaches have been developed for simplicity of use or similarity to laboratory motion analysis, but they have not been extensively examined. The main objective of the study was to determine the accuracy and repeatability of two common single-pose calibrations (N-pose and T-pose) under different conditions of placement (self-placement and passive placement), as well as their similarity to laboratory analysis based on anatomical landmarks. A further aim of the study was to develop two additional single-pose calibrations (chair-pose and stool-pose) and determine their accuracy and repeatability. Postures and movements of 12 healthy participants were recorded simultaneously with a full-body IMU suit and an optoelectronic system as the criterion measure. Three repetitions of the T-pose and the N-pose were executed by self-placement and passive placement, and three repetitions of the chair-pose and stool-pose were also performed. Repeatability for each single-pose calibration showed an average intraclass correlation coefficient for all axes and joints between 0.90 and 0.94 and a standard error of measurement between 1.5 degrees and 2.1 degrees. The T-pose with passive placement is recommended to reduce longitudinal axis offset error and to increase similarity to laboratory motion analysis. Finally, the chair-pose obtained the least longitudinal axis offset error amongst the tested poses, which shows potential for IMU calibration. (C) 2017 Elsevier B.V. All rights reserved.