显微CT旋转步长对牛腰椎松质骨重建图像及数据的影响

The effect of micro-CT rotating step on reconstructed image and data of bovini lumbar cancellous bone

中文关键词:  显微CT  旋转步长  三维重建图像  骨形态计量数据  牛

英文关键词:Micro-CT  Rotating step  Three-dimensional reconstructive image  Bone morphometry  Bovini

中文摘要:

  【摘要】 目的：探讨显微CT对牛腰椎松质骨扫描时不同旋转步长对扫描时间、磁盘占用、重建图像质量及骨形态计量学数据的影响，为松质骨扫描设定旋转步长提供实验依据。方法：取牛同一腰椎的8块矩形离体松质骨，尺寸为4×2×2mm，依次设定旋转步长为0.25°、0.30°、0.40°、0.50°、0.75°、1.0°、1.5°、2.0°进行显微CT扫描，将步长0.25°设为对照组，比较其他7个步长组的三维重建图像质量、扫描时间、占用磁盘空间情况；分析骨体积（bone volume，BV）、骨表面积（bone surface，BS）、骨表面积密度（bone surface/volume ratio，BS/BV）、板状骨小梁厚度[trabecular thickness（plate model），Tb.Th（pl）]、板状骨小梁分离度[trabecular separation（plate model），Tb.Sp（pl）]、板状骨小梁数量[trabecular number （plate model），Tb.N（pl）]、杆状骨小梁直径[trabecular diameter（rod model），Tb.Dm（rd）]、杆状骨小梁分离度[trabecular separation （rod model），Tb.Sp（rd）]、杆状骨小梁数量[trabecular number（rod model），Tb.N（rd）]、骨小梁模式因子（trabecular bone pattern factor，TBPf）共10项骨形态学参数的统计学差异。结果：与步长0.25°组比较：（1）步长0.30°~1.5°的6组三维重建图像质量有不同程度的降低，但骨小梁结构清晰可辨，仍可较好反映松质骨标本的解剖学特征，而步长2.0°组骨小梁结构模糊，图像质量明显下降；（2）随着旋转步长的增大扫描时间、磁盘占用均呈线性降低；（3）骨形态计量学10项数据分析表明，步长0.30°、0.40°组各项数据与步长0.25°组比较均无统计学差异（P>0.05），BV值自步长2.0°组出现统计学差异（P<0.05），Tb.Sp（rd）、TBPf值自步长0.50°组出现统计学差异（P<0.05），BS、BS/BV、Tb.Th（pl）、Tb.Sp（pl）、Tb.N（pl）、Tb.Dm（rd）、Tb.N（rd）7组数据自步长0.75°组出现统计学差异（P<0.05）。结论：对牛腰椎松质骨进行显微CT扫描时，旋转步长0.4°扫描时，在保证重建图像清晰、数据精确的前提下，扫描时间最短、磁盘空间占用最低，适用于精确分析椎体骨形态计量学三维数据；1.5°扫描时，虽数据结果存在误差，但扫描时间快，磁盘占用低，且重建图像清晰、骨小梁完整，适用于快速观测椎体骨小梁影像学信息；2.0°扫描时，图像模糊、数据误差显著，不推荐选用。

英文摘要:

  【Abstract】 Objectives: To explore the effects of rotation steps on the scanning time, disc occupation and the image reconstruction of micro computed tomography(micro-CT) under the same scanning distinguishability and to investigate the extent of this effect on the data of bone morphometry after image reconstruction. Methods: 8 fresh isolated lumbar cancellous bones from bovini with the same size were selected, the rotation steps of 0.25°, 0.30°, 0.40°, 0.50°, 0.75°, 1.0°, 1.5°, 2.0° were chosen under the same scanning distinguishability, 0.25° was used as reference control and compared with the other rotation steps with respect to 3D images of bone trabecula, scanning time and disc occupation, statistical analysis on bone morphology. The parameter data included: bone volume(BV), bone surface(BS), trabecular thickness(plate model)[Tb.Th(pl)], trabecular separation(plate model)[Tb.Sp(pl)], trabecular number(plate model) [Tb.N(pl)], trabecular diameter(rod model)[Tb.Dm(rd)], trabecular separation(rod model)[Tb.Sp(rd)], trabecular number(rod model)[Tb.N(rd)], trabecular bone pattern factor(TBPf). Results: Rotation step of 0.25° was the control group. (1)The image quality was decreased from 0.30° to 1.5°, but the structure of trabecula was still well captured, and the characters of cancellous bone were also clearly observed. When the rotation step increased to 2.0°, the structure of the trabecula could not be clearly captured. (2)With the increasing of rotating steps, scanning time and disc occupation showed linear decline. (3)The data of bone morphometry indicated as following: no significant differences were noted between 0.30° and 0.25°(P>0.05), or between 0.40° and 0.25°(P>0.05); but for all results: significant difference was found for BV from 2.0°(P<0.05), for Tb.Sp(rd) and TBPf from 0.50°(P<0.05), and for BS, BS/BV, Tb.Th(pl), Tb.Sp(pl), Tb.N(pl), Tb.Dm(rd) and Tb.N(rd) from 0.75°(P<0.05). Conclusions: Rotation steps of 0.4° can be applied for accurate analysis of three-dimensional data of bone morphometry due to its clear reconstruction image, accurate data, less time of sample scanning and less of disc occupation; rotation steps of 1.5° can be used for fast observation of bone trabecular imaging due to its less time of scanning and less disc occupation despite of the data error; while rotation steps of 2.0° is not recommended due to its indistinct and error image.