Three-Dimensional Rotational X-ray Imaging for Spine Surgery

Abstract

A validation study was done in which reconstructed three-dimensional rotational x-ray images were quantitatively compared with corresponding anatomic sections. To assess the accuracy of reconstructed images acquired on a three-dimensional rotational x-ray imaging device. Minimally invasive procedures have proven quite successful as alternatives for a multitude of open treatments. An unfavorable property of this type of procedure is a lack of direct vision of the operating area. Three-dimensional rotational x-ray imaging may be able to merge the advantages of computed tomography and fluoroscopy: real-time two-dimensional projections for fast visual feedback and three-dimensional reconstructions for detailed volumetric imaging of complex anatomy. Twenty traumatic thoracolumbar burst fractures were created and underwent pedicle screw instrumentation and balloon vertebroplasty. Subsequently, a three-dimensional dataset was obtained, and the midsagittal image was reconstructed. The specimens were sliced, and photographs were obtained. Multiple parameters on the reconstructed images and photographs were measured two times by two observers using a graphical method. The differences and standard deviations were calculated for the corresponding parameters and for the intraobserver data. The mean difference between the corresponding values ranged between -1.1 and 2.1 mm for all parameters. The standard deviation for the differences per parameter ranged between 1.2 and 3.2 mm. The intraobserver differences ranged from -0.8 to 1.4 mm, and the standard deviation varied between 0.4 and 2.4 mm. The reconstructed midsagittal images were accurate in all measured parameters. The three-dimensional rotational x-ray technique may prove to be valuable for less invasive spine surgery.