미세 컴퓨터 단층촬영에서 영상재구성 체적소 두께가 하악 골소주 분석에 미치는 영향

Abstract

The purpose of this study is to report the effect of increased reconstruction voxel size on morphological parameters of the three dimensional microstructure for manbibular trabecular bone which were measured with 3D microcomputed tomography.

Recently, many researchers have emphasized the importance of bone architectures as well as bone density. This study was also performed to describe three dimensional bone architectures.

In this study, human mandibles were used. The specimens were scanned and reconstructed into a 3D structure using a microcomputed tomography system(Skyscan 1076, Belgium).

The scanning voxel size used to create 2D images was 18μm. From the 2D images, 3D structural images were created with a voxel 18×18×18μm in a reconstruction voxel size which was ranged from 18μm to 180μm at intervals of 18μm. Eight morphological parameters were measured in reconstruction voxel size from 18μm to 180μm and each value was compared to the value measured in 18μm reconstruction voxel size. Therefore, this study is for investigating the largest reconstruction voxel size which does not change the values of morphological parameters for trabecular bone. The morphological parameters measured are bone volume(BV), bone volume fraction(BV/TV), bone surface(BS), bone surface/bone volume(BS/BV), structure model index(SMI), trabecular thickness(Th.Th), trabecular number(Tb.N), and trabecular seperation(Tb.Sp). The specimens were divided into two types of trabecular bone, and one is condyle and the other is mandubular body group. The trabecular architecture of the former is more even and regular than that of the latter. Fifteen volume of interests(VOI) were created in each group and the total of VOI was thirty. The mean value and standard deviation of each parameter in each reconstruction voxel size were calculated. Repeated ANOVA test was used to determine whether significant differences in the values from each parameter existed among the regions(at 95% confidence interval level, p<0.05) and the following results were drawn.

The differences in the values from bone volume(BV) and bone volume fraction(BV/TV) in the mandibular body group were significant in all reconstruction voxel sizes(p<0.05), but in condyle group were not significant in 36μm reconstruction voxel sizes(p>0.05).

The differences in the values from bone surface(BS) in the mandibular body group were significant in all reconstruction voxel sizes(p<0.05), but in condyle group were not significant in 36μm, 54μm, 108μm, 126μm and 144μm reconstruction voxel sizes(p>0.05).

The differences in the values from bone surface/bone volume(BS/BV) in the mandibular body group were significant in all reconstruction voxel sizes(p<0.05), but in condyle group were not significant in 36μm and 108μm reconstruction voxel sizes(p>0.05).

The differences in the values from structure model index(SMI) in the mandibular body group were significant in all reconstruction voxel sizes(p<0.05), but in condyle group were not significant in 36μm reconstruction voxel size(p>0.05).

The differences in the values from trabecular thickness(Tb.Th), trabecular number(Tb.N) and trabecular seperation(Tb.Sp) in both condyle and mandibular body group were significant in all reconstruction voxel sizes(p<0.05). In summary, for condyle group, 36μm was determined as the reconstruction voxel size in which the differences in the values from most parameters (except for trabecular thickness, trabecular number, and trabecular seperation) were not significant. However, for mandible body group, the differences in the values from all parameters were significant. Therefore, it seems to be hard to determine that 36μm is a reconstruction voxel size which does not change values in all parameters for both groups. Especially, a small reconstruction voxel size like 18μm was needed for the accurate imaging for trabecular thickness, trabecular number and trabecular seperation.