ABSTRACT: (1) Background: Although radiography performed on the subject in an upright position is considered the standard method for assessing sagittal cervical alignment, it is frequently determined, or reported, based on MRI performed on the subject in a supine position. (2) Methods: Cervical alignment observed in both imaging modalities was assessed using four methods: the C2-7 Cobb angle, the absolute rotation angle (ARA), Borden's method, and the sagittal vertical axis (SVA). Cervical alignment was determined (lordosis, kyphosis, and straight) based on radiography. Then, the diagnostic cut-off values for the MRI images and their corresponding diagnostic accuracies were assessed. (3) Results: The analysis included 142 outpatients. The determined diagnostic cut-off values for lordosis, using three measurements (Cobb angle, ARA, and Borden's method), were -8.5°, -12.5°, and 3.5 mm, respectively, and the cut-off values for kyphosis were -4.5°, 0.5°, and -1.5 mm, respectively. The cut-off value for SVA > 40 mm was 19.5 mm. The Cobb angle, ARA, and Borden's method, on MRI, showed high negative predictive values for determining kyphosis. The SVA on MRI measurements also showed high negative predictive values for determining >40 mm. (4) Conclusions: MRI measurements may be predictive of cervical alignment, especially for the exclusion of kyphosis and SVA > 40 mm. However, caution is needed in the other determinations using MRI, as their accuracies are limited.
