The computed tomography (CT) images were acquired to reconstruct 3-dimensional
(3D) models and coordinate systems of the knee for preoperative planning and
postoperative fluoroscopic study. The hip, knee, and ankle joints in the UKA
side were imaged under a 64-slice CT scan (Sensation 64; Siemens) for each
patient before surgery. The 3D surface models of the femoral head, femoral
condyles, tibial plateau, and ankle were reconstructed based on segmentation of
CT images using the region-growing method in Amira 6.7.0 (Thermo Fisher
Scientific). Patients underwent CT again 6 months after medial UKA surgery for
3D modeling of knees (both sides) and implanted prostheses using the same
protocol.
Before surgery, the anatomic bony landmarks on surface models of the hip center
and medial and lateral femoral epicondyles of the UKA knee were used to create a
femoral coordinate system, and the ankle center and medial and lateral tibial
plateau centers were used to create a tibial coordinate system, according to the
method of Grood and Suntay.
13
To ensure the consistency of coordinate definition, we mirrored and
aligned the 3D models of the preoperative UKA knee with the coordinate systems
of the native knee after using iterative closest points to construct the
coordinate systems.
5 ,37 (link)
The meshes of the knee models in the medial compartment were excluded in
the alignment procedure to reduce the effect of modeling error caused by CT
metal artifact in postoperative UKA knees. A 3D deviation analysis indicated
that the root mean square error of the alignment method was 0.28 ± 0.05 mm for
the femur and 0.32 ± 0.08 mm for the tibia on the UKA side and 0.44 ± 0.18 mm
for the femur and 0.47 ± 0.08 mm for the tibia on the native side. The
manufacturer provided 3D computer-aided design (CAD) models of femoral and
tibial implants that were used in motion measurement. Thus, the CAD models of
implants were aligned to reconstructed implants to determine the position of
implants relative to femoral and tibial coordinate systems. The root mean square
error of distances between the 3D CAD and reconstructed implant models was 0.27
± 0.06 mm for the femoral condyle and 0.33 ± 0.07 mm for the tibial
baseplate.
Zheng N., Dai H., Zou D., Dimitriou D., Wang Q, & Tsai T.Y. (2023). Altered In Vivo Knee Kinematics and Lateral Compartment Contact Position During the Single-Leg Lunge After Medial Unicompartmental Knee Arthroplasty. Orthopaedic Journal of Sports Medicine, 11(2), 23259671221150958.