RESUMO
The purpose of the study was to assess accuracy and repeatability of a noninvasive navigated (NIN) measurement of the lower limb axes by comparison with the conventional, invasive navigated (IN) technique. The tested hypothesis was that NIN measure will significantly differ from IN measure when performed on a supine patient under general anesthesia just prior to knee arthroplasty. The accuracy study involved 20 cases. The following measures were performed with both systems: maximal extension angle, coronal mechanical femoro-tibial angle at the previously assessed maximal extension angle, coronal mechanical femoro-tibial angle at 30° of knee flexion. NIN and IN measures were compared with the appropriate statistical tests. The repeatability study involved 14 cases. The same measures were performed with the NIN system twice by two independent observers. The intra- and inter-observer variability was assessed by the calculation of the intra-class correlation coefficient. After correction for the systematic biases, the differences between the two systems were not significant. A good correlation, a good coherence and an excellent agreement between NIN and IN measures of maximal extension angle and coronal femoro-tibial mechanical angle at maximal extension. But measures at 30° of flexion were inconsistent. The NIN system can be considered as an accurate and precise tool for the assessment of the knee extension angle and the coronal deformation at maximal extension prior to knee arthroplasty. But this system is less accurate and less repeatable when measuring coronal femoro-tibial mechanical angle at 30° of flexion and should not be used for this purpose.