Re-positioning effects of a full torso imaging system for the assessment of scoliosis

The effect of re-positioning (sway and breathing) on the reproduction accuracy of a low-cost torso imaging system for monitoring idiopathic scoliosis is assessed. The system utilizes a rotating positioning frame and a 3D digitizer. Four images taken at 90 degree intervals are needed to generate a torso image. Five volunteers having no scoliosis were used for the study. Four torso images of each volunteer were obtained. Ten evenly-spaced cross-sections are computed for each torso image. The distance between the centroids of successive cross-sections and a reference vertical axis are computed at each cross-section. The standard deviation of this distance at each cross-section is indicative of the variability due to sway. The variability due to breathing is assessed by computing the standard deviation of the widths and lengths of each cross-section for each volunteer. Results obtained indicate that the torso imaging system is adequate for assessing idiopathic scoliosis.