Abstract
Introduction
Poor clinical outcomes following total knee arthroplasty (TKA) can be related to improper alignment of the components. The main challenge is the variability in biomechanical references, especially in cases of severe deformity or dysplasia, and in determining the surgical landmarks intraoperatively. An intraoperative imaging tool can be very useful to assess the alignments when there is still a chance for correction. We investigated, on cadaveric specimens, the accuracy of using iso-centric (ISO-C) imaging (that reconstructs 3D from multiple 2D fluoroscopic images) for this purpose.
Methods
Six fresh frozen cadaveric knees were implanted with a standard TKA system and imaged using an ISO-C 3D C-arm (Arcadis Orbic ISO-C). Each knee was scanned two times with the Iso-C scanner and with appropriate image settings to capture the transepicondylar axis (TEA) and the tibial tubercle individually. A CT scan of each specimen was acquired as the reference for comparison.
The ISO-C 3D reconstructed volumes were analyzed on the C-arm. For the CT images, the 3D data were processed in Analyze software with the same objective. The surgical and clinical TEA was determined by moving and rotating an oblique cutting plane (Figure 1a:CT and 1c:ISO-C). This oblique slice was then moved distally to picture the femoral pegs (Figure 1b:CT and 1d:ISO-C). The angle between these two references (angle α in Figure 1) defined the rotational alignment.
For the tibial component, the first cutting slice was oriented parallel to the component. A second slice was defined just distal to the component, and then moved distally to find the tibial tubercle in the third slice. The orientation of the tibial component was determined by fitting a rectangular box to the component boundary (Figure 2a:CT and 2d:ISO-C). The bone orientation was defined by a line connecting the centroid of a polygon drawn over the boundary of the cortical bone (Figure 2b:CT and 2e:ISO-C) to the medial third of the tibial tubercle (Figure 2c:CT and 2f:ISO-C). Measurements were repeated five times, the overall accuracies determined in comparison to CT, and the correlation between the ISO-C and CT determined by the Spearman rank (P<0.05).
Results
Correlation between the ISO-C and CT measures of the femoral and tibial alignments was statistically significant (P=0.005 and P=0.018) with corresponding correlation coefficients of 0.94 and 0.89 (Figure 3). The overall accuracies calculated for all of the specimens were 0.3°± 0.8° for the femoral component, and 0°±1.4° for the tibial component. The calculated effective doses for the ISO-C imaging protocol of the femoral and tibial components were 0.005 mSv and 0.025 mSv respectively.
Discussion
This study showed that it is feasible to use ISO-C imaging for assessing the alignment of TKA components with acceptable accuracy both intraoperatively and postoperatively. The intraoperative assessment of ISO-C can help improve the outcome of knee arthroplasty and avoid early revisions because of complications related to component alignments. Results showed that it is also possible to use the ISO-C imaging as a safer modality (i.e. less radiation dose) for investigating links between proper component alignment and post-arthroplasty complications.
