Two-dimensional representation of three-dimensional pelvic motion during human walking: an example of how projections can be misleading.

We investigated the movements of skin markers located on the anterior and posterior surfaces of the pelvis during normal walking. Plots of the vertical versus horizontal displacements of the body-surface markers for a single gait cycle yield characteristic horizontal figure-of-eight patterns, often referred to as Lissajous figures. Some literature citations indicate that these figure-of-eight plots represent movement of the body center of mass in the plane perpendicular to the line of progression. We show evidence suggesting that the Lissajous plot for the body center of mass is U-shaped and that the observed figure-of-eight pattern is due to the location of the marker on the body surface coupled with pelvic rotation. A simple rigid-body model is used to demonstrate that pelvic rotation about the vertical axis can appear as horizontal translations in a planar projection. Even small rotations about the vertical axis are observable in the phase relation between horizontal and vertical displacements of surface markers in the projection. As a result, Lissajous plots of vertical versus horizontal displacements, particularly for points on the exterior of a rigid body, may be strongly influenced by rotations. We demonstrate that Lissajous patterns for the model are similar to patterns for the human pelvis during walking and that pelvic rotation has a large influence on Lissajous patterns (vertical movement vs medial-lateral) plotted from pelvic marker data. This demonstration illustrates how two-dimensional representations of three-dimensional movement can lead to incorrect interpretations.