Influence of Tomographic Slice Thickness and Field of View Variation on the Reproduction of Thin Bone Structures for Rapid Prototyping Purposes —An <i>in</i> <i>Vitro</i> Study

Abstract

This study assessed the influence of acquisition parameters of tomographic volumes on the reproduction of thin bone structures for rapid prototyping purposes. Two parameters were investigated: Field of View (FOV) and Slice Thickness (ST). The specimen was comprised of five pairs of 0.6 mm, 1.1 mm, 1.5 mm, 2.0 mm and 2.8 mm thick cortical bone plates. The plates were stuck into utility wax; the first plate of the pair was in vertical position while the second plate was oblique to the first one. Forty-five tomographic images were captured and separated into 3 groups of fifteen images. Each group had a specific FOV: 180 mm; 250 mm and 430 mm, respectively. Within each of these three groups, tomographic slice thickness was varied for every five of the fifteen slices. Acquisitions were carried out with STs of 1 mm, 2.5 mm and 5 mm. The Cyclops Medical Station software was used in the voxel-to-voxel analysis of radiologic density, reaching a total of 1350 assessed images. ST and FOV variation influenced the reproduction of thin bone walls, and FOV was shown to be a very important parameter. The larger the acquisition FOV, the more reduction in the number of voxels within the range of reconstruction for cortical bone in all of the bone plates. The visual analysis of the images of very thin bone walls showed that there could be a sharp drop in the radiologic density value in several adjacent voxels, resulting in areas which might not be reproduced in the reconstruction.

A New Level-Set-Based Protocol for Accurate Bone Segmentation From CT Imaging

“…The results show that segmentation accuracy is more dependent on the reconstruction FOV than on the slice thickness. This observation in in agreement with observations found in other phantom studies [12] [47]. Using the high-resolution point cloud to generate the surface meshes may allow the acquisition of image data with more clinical settings, reducing the exposure of the patient to the radiation without a significant loss in model accuracy.…”

“…Image reconstruction with large slice thicknesses is commonly associated with larger model inaccuracies [10][45] [46]. Nevertheless, in a recent study [47] it was demonstrated that alongside with the slice thickness, the FOV also influences the spatial resolution of the scan and the amount of partial volume averaging. The size of the voxel may be more critical than other potential error sources, such as the surface meshing parameters.…”

The effects of computed tomography scanner parameters on the quality of the reverse triangular surface model of the fibula

“…Moreover, the effects of variation in spacing between slices for the CT data acquired at a particular slice thickness and variation in other parameters are not considered. Meurer et al [19] concluded that the field of view had qualitatively more influence on the quality of 3D reconstructions than slice thickness for thin bone structures. Whyms et al [20] stated that the linear, angular, and volume measurements can be confidently obtained over a range of parameters for the mandible 3D CT models.…”

“…In addition, the accuracy of the 3D models depends on the geometry of the object [11,19]. However, the reported investigations [11,12,[19][20][21][22] mainly focus on the effects of slice thickness and spacing on the accuracy of the resulting surface models. Only a few studies [19][20][21][22] have been reported that consider other factors, but in a limited manner.…”