Dinko Nizić1. 1. Clinical Institute of Diagnostic and Interventional Radiology, Zagreb University Hospital Center, Kišpatićeva 12, 10000, Zagreb, Croatia. dinko.nizic@gmail.com
Abstract
OBJECTIVE: To compare the positions of the trochlear groove (TG) line and the vertical midline of the pericondylar rectangle on axial computed tomography (CT). MATERIALS AND METHODS: A retrospective pilot study was conducted in February and March 2011 comprising 14 knees in 9 consecutive patients with patellofemoral pain, a normal trochlear angle (124°-145°), and no medical record of prior knee surgery. After the axial CT image displaying the center of the trochlear groove ("reference cut") with a normal trochlear angle was identified, the pericondylar rectangle was drawn and located in the first quadrant of the Cartesian rectangular coordinate system. The x-coordinates of the vertical lines passing through the most posterior point of the trochlear groove (TG line) and the midpoints of the horizontal sides of the pericondylar rectangle (vertical midline) were obtained at 2-week intervals and statistically analyzed as matched pairs for differences (Wilcoxon signed rank test), agreements (Bland-Altman plot, intraclass correlation coefficient [ICC]) and relative variations (coefficient of variation [CV]). RESULTS: There were no statistically significant differences between matched pairs (2-tailed p from 0.583 to 0.641) whereas the agreements were substantial (bias = -0.37 and -0.45 respectively, ICC = 0.688 and 0.670 respectively) to almost perfect for duplicate measurements (bias = 0.11 and 0.04 respectively, ICC = 0.975 and 0.998 respectively), with much less relative variation with regard to the vertical midline (CV = 1.22% to CV = 0.34% respectively). CONCLUSION: The positions of the TG line and the vertical midline of the pericondylar rectangle on axial CT were identical. The latter was identified regardless of trochlear morphology.
OBJECTIVE: To compare the positions of the trochlear groove (TG) line and the vertical midline of the pericondylar rectangle on axial computed tomography (CT). MATERIALS AND METHODS: A retrospective pilot study was conducted in February and March 2011 comprising 14 knees in 9 consecutive patients with patellofemoral pain, a normal trochlear angle (124°-145°), and no medical record of prior knee surgery. After the axial CT image displaying the center of the trochlear groove ("reference cut") with a normal trochlear angle was identified, the pericondylar rectangle was drawn and located in the first quadrant of the Cartesian rectangular coordinate system. The x-coordinates of the vertical lines passing through the most posterior point of the trochlear groove (TG line) and the midpoints of the horizontal sides of the pericondylar rectangle (vertical midline) were obtained at 2-week intervals and statistically analyzed as matched pairs for differences (Wilcoxon signed rank test), agreements (Bland-Altman plot, intraclass correlation coefficient [ICC]) and relative variations (coefficient of variation [CV]). RESULTS: There were no statistically significant differences between matched pairs (2-tailed p from 0.583 to 0.641) whereas the agreements were substantial (bias = -0.37 and -0.45 respectively, ICC = 0.688 and 0.670 respectively) to almost perfect for duplicate measurements (bias = 0.11 and 0.04 respectively, ICC = 0.975 and 0.998 respectively), with much less relative variation with regard to the vertical midline (CV = 1.22% to CV = 0.34% respectively). CONCLUSION: The positions of the TG line and the vertical midline of the pericondylar rectangle on axial CT were identical. The latter was identified regardless of trochlear morphology.