The presented approach was evaluated for accuracy by a phantom study. Awooden stick was used to simulate a femoral shaft. Smaller sticks penetrating the ends of this "shaft" simulated two condylar axes and three femoral neck axes. The arrangement is depicted in Fig. 15. The angles between the axes were measured by conventional goniometers. The phantom was then placed in a CT scanner in an arbitrary position comparable to a patient's femur and the scanning protocol FIGURE 13 Determination of the pelvis axis II (left image, anterior described above was applied. The scan was then repeated with
the phantom placed in a different position. Digital images were
segmented and the angles of the 3D reconstruction were determined for the two scans, resulting in 6 shaft-neck angles and 12 torsional measurements of the neck. The measurements of the 3D digital model are compared with the measurements on the phantom. The average difference ( + one standard deviation) for the shaft-neck angles is —0.02° + 0.84°, and for the torsion measurement +0.06° + 1.16°.
Was this article helpful?