Measuring small linear displacements with a three-dimensional video motionanalysis system: Determining its accuracy and precision

Objectives: To determine accuracy, precision, and smallest detectable diffe rence for a three-dimensional (3D) video motion analysis system specificall y configured for measuring small and slow displacements within a small meas urement volume (0.7 x 0.5 x 0.3m). Design: Repeated measurements with random sequence of conditions. Setting: Rehabilitation research laboratory Intervention: A reference sliding device was used to control cyclic displac ements of two reflective markers over 5 calibrated ranges (1, 3, 10, 30, an d 60mm). Nine cycles were performed for each of 9 conditions (3 directions X 3 zones). Four cameras recorded all trials on tapes, which were digitized with a Kinemetrix system. Main Outcome Measures: Change in distance of the moving markers relative to a third static marker was averaged over 50 frames per trial. Mean error, m ean absolute error, and intertrial and intratrial standard deviations (SDs) were calculated for each zone and direction. Results: For 810 trials, mean error and absolute error were, respectively,. 034mm and .094 mm. The mean intertrial and intratrial SDs and 99% confidenc e interval were .047mm (CI = +/- .121mm) and .030mm (CI = +/- .077mm). The corresponding smallest detectable differences were .171 mm and .109 mm. Conclusion: Motion analysis configured for registration within small volume s allows measurement of minuscule displacements with great accuracy and may therefore be suitable for many applications in rehabilitation research oth er than gait analysis. (C) 1999 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation.