Technical Considerations and Accuracy Improvement of Accelerometer-Based Portable Computer Navigation for Performing Distal Femoral Resection in Total Knee Arthroplasty.

<bold>Background: </bold>Accelerometer-based computer navigation has been shown to be highly accurate for performing distal femoral and proximal tibial component alignment in total knee arthroplasty (TKA), although the procedure for the femoral component is less accurate than for the tibial component.<bold>Methods: </bold>First, 30 knees without hip osteoarthritis or proximal femoral surgeries were selected. Sequential hip adduction, abduction, and flexion were performed, and the femoral head was monitored fluoroscopically in the coronal plane before TKA. Significantly more movement was detected during hip adduction than during abduction and flexion. Then, postoperative femoral and tibial component alignment was retrospectively evaluated in 48 TKAs before fluoroscopic monitoring (early group) and in the next 61 TKAs with femoral registration using smaller adduction movements to avoid large femoral head movements (later group). Another 47 TKAs treated with the conventional intramedullary method for the distal femoral component and the extramedullary method for the proximal tibial component were also analyzed (IM and EM group) for historic control.<bold>Results: </bold>Significantly large variances in the femoral component implantation of the early group were detected in both the coronal and sagittal planes. The sagittal femoral implantation angle of the early group (4.6 ± 3.0°) was significantly larger than that of the later group (3.2 ± 1.8°) when 3.5° was the target for both groups. No significant difference was detected in the variances of either the coronal or sagittal tibial component implantation, although the coronal tibial implantation angle was significantly smaller (-1.3 ± 1.3°valgus) in the early group than in the other groups.<bold>Conclusion: </bold>Accelerometer-based navigation sometimes has technical issues during registration associated with hip adduction. We showed that femoral registration without large adduction movements will enable more accurate femoral implantation. Surgeons should also keep in mind that the coronal tibial component is likely to be in valgus alignment (about 1°) even if a neutral angle (0°) is selected with this particular device. [ABSTRACT FROM AUTHOR]