Abstract
INTRODUCTION
The outcome after total knee arthroplasty is influenced by the postoperative orientation of the component. For accurate implantation, the surgeon performs a three dimensional preoperative planning and performs the surgery with reference to the anatomical bony landmarks. However, the assessment of orientation after TKA is generally performed on two dimensional radiographs. Despite the accurate implantation, radiographic assessment may not able to accurately evaluate the orientation of the component. CT images obtain a three dimensional information after TKA, but reliable identification of the anatomical bony landmarks remains the problem due to artifacts of metal components. In this study, we evaluate the three dimensional orientation of the component relative to the bone axis of anatomical landmarks using pre- and post-operative CT scanning.
PATIENTS AND METHODS
Two knees after primary TKA were assessed by one observer using preoperative and postoperative CT images. 3D models of pre-operative bone and post-operative bone with the exclusion of component data were constructed. Surface-based registration was performed by independently implementing the iterative closest point algorithm with the least-squares method to match the pre-operative bone model with the post-operative bone model. 3D surface model of the metal component from postoperative CT images was constructed. 3D surface model of the metal component was superimposed on original computer-aided design (CAD) data of the component using surface-based registration. The registration of the metal component was performed three times. Intra-observer reliability of the superimposed CAD models was evaluated. The orientation of the component was measured in euler angle between the axis of the superimposed CAD model and the bone axis of anatomical landmark.
RESULT
The root-mean-square (RMS) errors for femoral component were 0.14°, 0.27°, and 0.64° for rotation in the coronal, axial, and sagittal planes, respectively. The RMS errors for tibial component were 0.06°, 0.39°, and 0.08° for rotation in the coronal, axial, and sagittal planes, respectively. The orientation of the femoral component relative to the axis of anatomical bony landmark was 1.3° of varus, 3.7° of internal rotation, and 7.4° of extension. The orientation of tibial component was 1.5 ° of varus, 19.5° of internal rotation, and 5.2° of flexion.
DISCUSSION AND CONCLUSION
The postoperative orientation of component is important for long-term survival after TKA. In this study, the RMS errors were 0.7° or less in axial plane in the coronal, axial, and sagittal planes. This technique is considered useful to evaluate the three dimensional orientation of the component after TKA.
