Purpose: To test the diagnostic performance of susceptibility-weighted MRI (sMRI) for the evaluation of vertebral body fractures versus standard MRI-sequences, using CT as reference standard.
Methods: In this prospective study 88 vertebral fractures (45 healed, 43 non-healed) were detected in 39 patients who underwent T1/T2/TIRM MRI-sequences and sMRI. All fractures were evaluated with CT as reference standard. In all modalities/sequences, displacement and height of the posterior vertebral body cortex and visibility of fracture lines and cortical breaks were assessed. Sensitivity, specificity and inter-reader agreement between MRI and CT were calculated.
Results: sMRI demonstrated highest diagnostic accuracy for detection of posterior vertebral body cortex involvement (sensitivity: 98 %/specificity: 100 %), fracture lines (86 %/99 %) and cortical breaks (93 %/100 %) versus T1/T2/TIRM sequences. Regarding evaluation of posterior vertebral body cortex displacement and height, sMRI demonstrated the closest intermodality agreement (R2=0.96; 95 % CI -0.92-0.89/R2=0.97; 95 % CI -1.67-1.23) with CT and the closest interobserver agreement (R2=0.97; 95 % CI -0.71-1.01).
Conclusion: sMRI allows reliable evaluation of vertebral body fractures with regard to posterior vertebral body cortex displacement and height, cortical breaks and fracture lines with higher accuracy versus standard MRI, especially in patients with non-healed vertebral body fractures.
Key points: • sMRI allows a reliable evaluation of vertebral body fractures. • sMRI has higher accuracy than standard-MRI for evaluation of vertebral body fractures. • sMRI is especially useful in patients with non-healed vertebral body fractures.
Keywords: Computed tomography; MRI; Musculoskeletal; Susceptibility-weighted magnetic resonance imaging; Vertebral body fractures.