Robert Kerberger^1,*; Bert van Rietbergen²; Dieter Drescher¹; Kathrin Becker¹

¹Department for Orthodontics, University Hospital Duesseldorf, Duesseldorf, Germany; ²Department for Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands

Background: Micro-CT enables volumetric analysis of peri-implant bone tissue. However, it is not yet clear to what extent beam hardening artifacts affect local grey value intensities depending on the distance to the implant.

Aim/Hypothesis: The present study aims at assessing (i) to what extent peri-implant grey values are affected and (ii) whether the values can be corrected mathematically.

Material and Methods: Two customized grade V titanium mini implants were placed in the dorsal portion of a rat tail vertebra in nine female Wistar rats. After scarification, the animals were scanned with a µCT. Then, the implants were carefully removed, and another µCT scan was performed (reference scan). The scans were superimposed, and the grey value differences were assessed in a standardized volume of interest (VOI) at corresponding voxels and increasing distances to the implant (from 10.4 µm in 12 steps to 405.6 µm). A correction coefficient (CC) was finally estimated by means of piecewise polynomial function (PPF), considering the distance to the implant as independent and the grey-value difference as dependent variable. To validate the effectiveness of the CC, the amount of calcified bone volume per tissue volume (BV/TV) were assessed in VOI before and after application of CC, and effectiveness was compared by means of Wilcoxon signed rank test.

Results: At increasing distances, the discrepancy of gray values saturated. At 10.4 µm, the agreement amounted to 32.7% +- 6.11%, whereas at 83 µm distance, it amounted to 79.8%+-5.35% and at 405.6 µm, it amounted to 93.4% +- 4.46%. The goodness of fit amounted to R2=0.989, RMSE=0.0305. The Wilcoxon signed rank test confirmed significant improvement of gray values (p < 0.001). Nonetheless, when splitting the VOI into segments, overcorrection was sometimes achieved in proximity to the implant.

Conclusion and Clinical implications: Within the limits of the study, the present investigation confirmed that gray values are more affected by beam hardening in close proximity to the implant. A logarithmic relationship between distance to implant and grey value difference was found. Correction of gray value was possible with a small residual error by means of a piecewise polynomial function.

Disclosure of Interest: None Declared

Keywords: animal study, micro-CT, orthodontics

罗伯特·克伯格^1,* ; 伯特范里特伯根² ; 迪特·德雷舍¹ ; 凯瑟琳贝克¹

¹德国杜塞尔多夫杜塞尔多夫大学医院正畸科；²埃因霍温理工大学生物医学工程系，荷兰埃因霍温

背景： Micro-CT 能够对种植体周围骨组织进行体积分析。然而，目前尚不清楚光束硬化伪影在多大程度上影响局部灰度值强度取决于与植入物的距离。

目的/假设：本研究旨在评估 (i) 种植体周围灰度值的影响程度以及 (ii) 是否可以通过数学方式更正这些值。

材料与方法：两个定制的 V 级钛微型植入物被放置在 9 只雌性 Wistar 大鼠的大鼠尾椎骨的背部。划痕后，用 μCT 扫描动物。然后，小心地移除植入物，并进行另一次 μCT 扫描（参考扫描）。扫描被叠加，并在相应体素的标准化感兴趣体积 (VOI) 中评估灰度值差异，并增加与植入物的距离（从 10.4 µm 分 12 步增加到 405.6 µm）。最终通过分段多项式函数 (PPF) 估计校正系数 (CC)，考虑到植入物的距离为独立变量，灰度值差异为因变量。为了验证 CC 的有效性，

结果：随着距离的增加，灰度值的差异饱和。在 10.4 µm 处，一致性达到 32.7% +- 6.11%，而在 83 µm 距离处，达到 79.8%+-5.35%，在 405.6 µm 处，达到 93.4% +- 4.46%。拟合优度为 R2=0.989，RMSE=0.0305。Wilcoxon 符号秩检验证实了灰度值的显着改善 ( p < 0.001)。尽管如此，当将 VOI 分成多个段时，有时会在植入物附近实现过度矫正。

结论和临床意义：在研究范围内，本调查证实灰度值更受植入物附近光束硬化的影响。发现了到植入物的距离和灰度值差异之间的对数关系。通过分段多项式函数，可以用很小的残差校正灰度值。

利益披露：无申报

关键词：动物研究，显微CT，正畸