Benchmarking of a full-body inertial motion capture system for clinical gait analysis

In order for gait analysis to be established as part of routine clinical diagnoses, an accurate, flexible and user-friendly motion capture system is required. Commonly used optical, mechanical and acoustic systems offer acceptable accuracy and repeatability, but are often expensive and restricted to laboratory use. Inertial motion capture has seen great innovation in the last few years, but the technology is not yet considered mature enough for clinical gait analysis. In this paper we compare the kinematic reliability of inertial motion capture with optical motion capture during routine gait studies of eight able-bodied subjects. The root mean squared, RMS, and coefficient of correlation, R, was used to compare data sets. Saggital plane joint angles in the knee and hip compared very well. Corresponding transverse and frontal plane values were moderately accurate. The ankle joint angles calculated from the two systems were less accurate. This was believed to be due to the use of different rotation axis orientations used for calculation of angular rotations.