IntroductionThe spinal distortion that gives rise to a scoliotic curve includes both a rib hump and an axial rotation of the apical vertebra. The location and extent of axial rotation can provide considerable information on the nature of the deformity. The earliest method of detecting an axial rotation was from the lateral displacement of the shadow of the spinal process with respect to that of the vertebral body on coronal plain radiographs. Subsequently, Grossiord et al. [12] and Nash and Moe [21] showed that the position of the pedicle shadow with respect to those of the vertebral body wedges and the posterior arch provided a more sensitive index. They also showed that accurate calibration of the X-ray machine was most important, as the coronal view of the spine provides only a frontal projection of the vertebrae, and hence a qualitative evaluation of the axial rotation.Several authors [4, 5,7,8,19,20,24,32] have used a variety of geometrical formulae that take into account the diverse anatomical frames of the vertebrae as projected in coronal radiographs to develop a quantitative measurement of axial rotation, implicitly referred to the global referential system. Indeed, the concept of axial rotation of the scoliotic curve vertebrae requires a single referential system for all the vertebrae. The accuracy of these methods has been assessed [14,28,29,32,35] and the shadows of the anatomical features were found to be distorted [9, 25, 31 35]. These evaluations have also been compared to those drawn from CT scans and ultrasound images [1, 2,12,16,17,18,28], and it has been shown that the measurement technique must be standardized to allow the comparative assessment of findings [29].However, none of the above methods use sagittal data, consequently they do not produce a true three-dimensional (3D) measurement of the axial rotation of each vertebra. The 2D data provide what is referred to in this paper as "apparent rotation". True 3D measurements must also take into account the sagittal tilt of the vertebrae. This report analyses the concept of introduced rotation and its implications, and then describes a method of obtaining a true 3D evaluation of vertebral axial rotation.Abstract Scoliotic curvatures can only be assessed through three-dimensional (3D) procedures. Measurement of the axial vertebral rotation appears to be of primary importance for such techniques. Nevertheless, traditional methods are based only on 2D data, obtained through antero-posterior radiographic projections of the spine. A 3D method is described in this study, taking into account the sagittal tilt of the vertebrae. Only such a measurement provides a real 3D method for a true appraisal of the scoliotic spine. The practical implications are developed.