A Precise and Reliable Method of Determining Lesion Size in Osteonecrosis of the Femoral Head Using Volumes

Background To ascertain whether volumetric measurements to characterize lesion size in osteonecrosis of the femoral head using magnetic resonance imaging (MRI) and 3D software are more precise than other previously described methods. Methods Twenty-four patients were included in the study. Two musculoskeletal radiologists independently analyzed radiographs and MRIs using the methods described by Kerboul et al [9], Koo and Kim [10], and Cherian et al [11]. Volumetric measurements were calculated from the MRIs using 3D imaging software. Inter-rater reliability was calculated for all 4 methods using the interclass correlation coefficient (ICC). Levene's test was used to compare the variance across methods, serving as a measure of precision of each method. Results An ICC value of 0.81 was calculated for the volumetric measurements. The ICC values of the Kerboul et al, Koo and Kim, and Cherian et al methods were 0.94, 0.61, and 0.49, respectively. Levene's test for homogeneity of variance using absolute deviations showed the variance was not equal across methods (P < .01). The variance and the corresponding 95% confidence interval were calculated showing that the variance for the volumetric measurements was the smallest among the 4 methods examined, indicating that the volumetric measurements are more precise in characterizing lesion size as compared to the other methods. Conclusion Volumetric measurements of lesion size using 3D MRI imaging software to assess osteonecrosis of the femoral head are more precise than previously described methods and have excellent interobserver reliability. A 3D MRI assessment of volume of osteonecrosis in the femoral head may be useful in clinical decision-making.