Accuracy of in vitro mandibular volumetric measurements from CBCT of different voxel sizes with different segmentation threshold settings

Abstract

Background: To determine the accuracy of volumetric measurements of the mandible in vitro by cone-beam computed tomography (CBCT) and to analyze the influence of voxel sizes and segmentation threshold settings on it.

Methods: The samples were obtained from pig mandibles and scanned with 4 voxel sizes: .125 mm, .20 mm, .30 mm, and .40 mm. The minimum segmentation thresholds in Hounsfield units (HU) were set as 0, 100, 200, 300, and 400, respectively, for each voxel size for 3D reconstruction. Laser scanning as the reference, the volumes of each CBCT scanning, the mean iterative distances of superimposition and total positive and negative deviations were recorded and compared.

Results: The volumes of CBCT-scan deviated from those of laser-scan by + 7.67% to - 3.05% with different HU and voxel sizes. The deviation increased with the voxel size. There was a more suitable minimum HU threshold of segmentation (HU100 for .125 mm, 200 for .20 mm, 300 for .30 mm, and 400 for .40 mm) for each voxel size.

Conclusions: Voxel sizes and Hounsfield unit thresholds influence the accuracy of volumetric measurements in CBCT scanning. The volume increase with the voxel size, and different voxel sizes correspond to different optimal Hounsfield unit thresholds.

Keywords: Cone-beam computed tomography (CBCT); Hounsfield unit threshold; Volumetric measurement; Voxel size.

Fig. 1

CBCT scanning register with laser scanning. Each CBCT scanning file of different voxel size (.125 mm, .20 mm, .30 mm, .40 mm) and different HU (0, 100, 200, 300, 400) was then individually superimposed on the laser scanning file (labeled as the reference file) by means of an automated best-fit algorithm. A .5-mm threshold parameter was set as the critical value to analyze deviations between the laser scanning file (reference file) and each CBCT scanning file (test file). The darker the color is, the larger the variance is, and the lighter the color is, the smaller the variance is. For the same HU value (column), with the increase of voxel size, the color becomes darker, showing the increase of the variance between the CBCT file and the laser file. For the voxel size (line), each voxel size has an optimal HU value with the lightest color (100 for .125 mm, 200 for .2 mm, 300 for .3 mm, and 400 for .4 mm)

Fig. 2

Mean iterative distance of the superimposition. The mean iterative distance of the superimposition of each CBCT scanning file and laser scanning file was auto-calculated. For the .125-mm voxel size, the least iterative distance was achieved when we chose 100 as the minimum HU threshold of segmentation. This equally applied to 200 for the .2-mm voxel size, 300 for the .3-mm voxel size, and 400 for the .4-mm voxel size as the minimum HU threshold of segmentation. Statistical significance can be seen in the .4-mm group (P = .07)

Fig. 3

Mean total percentages of the points outside the bounds of the superimpositions. A 5-mm threshold parameter was set as the critical value to analyze deviations between the laser scanning file (reference file) and each CBCT scanning file (test file). Reports were generated for calculating the total positive and negative deviations separately. Seen as a whole, for the .125-mm voxel size, the lowest percentage was achieved when we chose 100 as the minimum HU threshold of segmentation. Similarly, 100 and 200 HU created the lowest percentage for the .20-mm voxel size, 200 and 300 HU created the lowest percentage for the .30-mm voxel size, and 300 and 400 HU created the lowest percentage for the .40-mm voxel size