How Does the Supine MRI Correlate With Standing Radiographs of Different Curve Severity in Adolescent Idiopathic Scoliosis?

Study Design. A retrospective study. Objective. To study how the supine magnetic resonance image (MRI) correlates with standing radiographs of different curve severity in adolescent idiopathic scoliosis (AIS). Summary of Background Data. Linear correlation between Cobb angles measured on supine MRI and standing radiographs has been identified. However, the effects of different curve severity on the correlation have not been studied in depth. Methods. Girls with AIS with standing radiographs and supine MRI were reviewed. From standing radiographs, all structural and nonstructural Cobb angles were measured. For those with simultaneous lateral radiographs, thoracic kyphosis (TK) and lumbar lordosis (LL) angles were measured. On supine MRI, the coronal Cobb angles, TK and LL were measured accordingly. The coronal Cobb angles were divided into 3 groups based on values measured on standing radiographs: mild group for Cobb angles less than 20 degrees, moderate group for 20 degrees to 40 degrees, and severe group for more than 40 degrees. Correlation was analyzed using scatter plot. Results. Eighty patients with AIS with 122 coronal curves were reviewed. On standing radiographs, the coronal Cobb angles were 14.7 degrees +/- 3.2 degrees, 28.2 degrees +/- 5.1 degrees, and 54.9 degrees +/- 11.3 degrees for mild, moderate, and severe groups. On supine MRI, the Cobb angles averaged 10.1 degrees +/- 5.6 degrees, 20.0 degrees +/- 6.3 degrees, and 49.4 +/- 12.3 degrees for each group, respectively. TK were 16.3 +/- 9.1 degrees and 11.8 +/- 6.1 degrees for radiographs and MRI (P < 0.001), whereas the LL averaged 45.5 +/- 12.2 degrees and 39.5 +/- 10.5 degrees for radiographs and MRI (P < 0.001). Cobb angles measured on standing radiographs and supine MRI were linearly correlated with the adjusted R-2 being 0.0627, 0.2118, and 0.7999 for the mild, moderate, and severe groups. Conclusion. Cobb angles measured on supine MRI were linearly correlated with Cobb angles measured on standing radiographs and the correlation was more reliable in those with Cobb angles more than 40 degrees. Therefore, the supine MRI could serve as a reliable alternative to standing radiographs in the assessment of Cobb angles more than 40 degrees in AIS.