Trochlear boss height measurement: A comparison of radiographs and MRI

Abstract

A key anatomical consideration and determinant of surgical approach in trochlear dysplasia is the trochlear boss height (TBH), traditionally defined by measurements on plain X-rays (XR). Magnetic resonance (MR) imaging is increasingly used for pre-operative planning and follow-up. However, it is unclear whether measurement of TBH on XR is applicable to MR. The aim of this study was to establish the reliability of TBH measurement on MR compared to XR. This study used lateral knee radiographs and MR scans of 14 knees of patients with trochlear dysplasia, six knees of non-dysplastic patients with anterior knee pain (AKP), and five knees of non-dysplastic controls with no AKP. Correlation between XR and MR measurements was assessed using Pearson correlation coefficients. Agreement between methods and observers was assessed using Bland-Altman plots with 95% limits of agreement. Intra- and inter-observer reliability was assessed using intraclass correlation coefficients (ICC). Bland-Altman charts showed a total width of 95% limits of agreement of 4.78 mm for XR and MR subchondral bone (SB) TBH measurements, and 6.73 mm for XR and MR cartilage TBH measurements. Inter-observer ICCs were 0.86 for XR, 0.62 for MR SB, and 0.53 for MR cartilage. The widths of the Bland-Altman 95% limits of agreement between observers were 4.79 mm (XR), 5.04 mm (MR SB) and 4.74 mm (MR cartilage). Measurement of TBH on MR is not directly interchangeable with XR. Adjustments need to be made to treatment thresholds based on XR measurement if MR is used instead.