Consideration of digitization precision when building local coordinate axes for a foot model

This study investigated whether points digitized for the purpose of embedding coordinate systems into the foot accurately represented the orientation of the bone described. Eight complete data sets were collected from 9 adult cadaver specimens. Palpable landmarks defined 5 segments to include the calcaneus, navicular, medial cuneiform, first metatarsal, and hallux. With use of the Flock of Birds electromagnetic motion tracking device, a single examiner digitized a minimum of 3 points for each segment. Coordinate definitions followed the right-hand rule, with left-sided data converted to right-sided equivalency. Local axes were created where X projected approximately forward, Y upward, and Z laterally. Matrix transformation computations calculated the angular precision in degrees between coordinates built from points digitized pre- and post-dissection of surface tissues covering bone. The condition of post-dissection was considered the criterion standard for comparison. Change about the X-axis represented the angular precision of the coordinate in the frontal anatomical plane; Y-axis in the transverse plane; Z-axis in the sagittal plane. The calcaneus and navicular coordinate axes changed by an average of <3 degrees across conditions. Mean coordinate angulation of the cuneiform X, Y, Z axes changed by 6.0 degrees , 4.6 degrees , 11.9 degrees , respectively. Change in coordinate angulation was largest for the X-axis at the first metatarsal (48.6 degrees ) and hallux (36.5 degrees ). A two-way repeated measures ANOVA found a significant interaction between the axis and segment (F=8.87, P=0.00). Tukey post-hoc comparisons indicated the change in coordinate angulation at the X-axis for the cuneiform, metatarsal, and hallux to be significantly different (P <0.05) from the calcaneus and navicular. The X-axis of the first metatarsal and hallux was different from all other axis-segment combinations except for the Z-axis of the cuneiform. Differences in locating landmarks reduced angular precision of the coordinate axes most in the smallest foot segments where points digitized were located close together. We can recommend the proposed landmarks for the calcaneus and navicular segments, but kinematics determined about the coordinate axes for the small sized medial cuneiform, and the long (X) axis for the first metatarsal and hallux have excessive error.