MRI-based static and functional assessment of complex hip deformities in comparison with CT: a validation study

This study aimed at investigating the agreement between predefined quantitative parameters of hip morphology derived from magnetic resonance imaging (MRI) and virtual range of motion (ROM) analysis using computed tomography (CT) as standard of reference.

Twenty patients (13 females, 7 males, 16-59 years) with hip deformities underwent MRI prior to surgery. Clockwise alpha angle, femoral head and neck diameter, collum caput diaphysis angle, femoral torsion, center-edge angles, acetabular coverage of the femoral head, sourcil angle, and acetabular anteversion were measured. Additionally, tern single and combined movements were simulated using a motion analysis program. The MRI findings were compared with the corresponding results obtained by CT. Correlation of MRI with CT was assessed using different statistical methods (intraclass correlation coefficient, Bland-Altmann plot, two one-sided t test), and linear regression analysis was performed.

The results showed near-perfect intraclass correlation coefficients (ICCs) for anteversion (0.95), acetabular sector angles (0.98-0.99), sourcil angle (0.95), and acetabular coverage (anterior 0.96, posterior 0.99). Intermethod correlation for femoral parameters showed almost perfect agreement except for the alpha angle (0.73-0.97). No significant proportional bias was detected for traditional acetabular and femoral parameters. ROM analysis was performed for 370 movements in 37 hips. For 78.4% (290/370) of the movements analysed, neither CT nor MRI detected impingement across the physiological ROM. For 18.6% (69/370) of the movements, impingement was detected by both CT and MRI, while 2.2% (8/370) of the movements with impingement in CT showed no impingement in MRI, and 0.8% (3/370) of the movements with impingement in MRI had no corresponding result in CT.

Finally, it was concluded that MRI-based assessment of hip morphology and virtual ROM analysis is feasible and can be performed with good intermethod agreement in comparison to the gold standard (CT). Therefore, MRI appears to be substantially equivalent to CT for use in virtual ROM analysis and so may reasonably be used in place of CT for this purpose.