Relative motion between tracking markers attached the leg and the underlying bones is the most significant source o f error in video based gait analyses. The purpose of this study was to assess the effect of the location, attachment method, and the physical characteristics of marker arrays over the lateral shank and the medial border of the tibia (MBT) on estimates of tibial rotations during the stance phase of free-speed walking. Seven subjects participated in this study. A root mean square deviation (RMSD) of the relative helical angle (RHA) was used to quantify rotational deviations between estimates made using eleven different surface marker arrays and the “true” orientation of the tibia as calculated from markers attached to a bone anchored percutaneous skeletal tracker (PST). The location of the marker arrays over the lateral shank was the only factor to statistically affect estimates of tibial rotation. Better estimates of tibial rotation were realized when the marker arrays over the lateral shank were placed more distal than proximal. Trends in the data suggest that better estimates of tibial rotation resulted when marker arrays over the lateral shank and MBT were attached using underwrap and overwrap methods respectively.

A second objective of this study was to identify the marker set that yielded the best estimate of tibial rotation. The physically constrained marker set over the distal lateral shank, attached using an underwrap method (DLU) was the best set of markers based on the ranking criteria used in this study. The RHA for the DLU marker set was decomposed about anatomically meaningful axes of the tibia. The majority of the rotational deviation occurred about the Z axis of the tibia during the first and last thirds of stance with relatively little error noted during mid-stance. Rotational deviations of ± 2 degrees about the X and Y axes, and ± 4 degrees about the Z axis did occur even when using the best set of markers.