Gyf Ng

Comparison between an accelerometer and a three-dimensional motion analysis system for the detection of movement

OBJECTIVE To examine the reliability and concurrent validity of an accelerometer compared with a three-dimensional (3D) motion analysis system for measuring the motor reaction time of the lower limbs, to validate a simple method for objective clinical assessment of movement onset. DESIGN Timing of the onset of knee extension movement in response to an audio signal was measured simultaneously with an accelerometer and a 3D VICON motion analysis system. PARTICIPANTS Twelve able-bodied subjects with a mean age of 26 (standard deviation 2.3) years. RESULTS Good reliability was found for both instruments, although the intraclass correlation coefficient (ICC) was higher for the accelerometer (ICC₃,₁=0.739; P<0.001). The mean motor reaction time measured by the accelerometer and the VICON system was 205.0 and 196.9 mseconds, respectively. Good agreement was found between the paired measurements (mean ICC=0.774), and the average 95% limits of agreement were -56.4 to 72.5 mseconds. CONCLUSION In conclusion, the accelerometer had good reliability for the measurement of movement onset time. Agreement between the measurements from the accelerometer and the VICON system was high. However, the limits of agreement covered a wide range, so absolute timings for movement onset derived from these methods should not be used interchangeably.