Institute of Medical Engineering, Graz University of Technology, Graz, Austria.
Comput Med Imaging Graph. 2012 Mar;36(2):85-94. doi: 10.1016/j.compmedimag.2011.07.006. Epub 2011 Sep 8.
One major source affecting the precision of bone structure analysis in quantitative magnetic resonance imaging (qMRI) is inter- and intraoperator variability, inherent in delineating and tracing regions of interest along longitudinal studies. In this paper an automated analysis tool, featuring bone marrow segmentation, region of interest generation, and characterization of cancellous bone of articular joints is presented. In evaluation studies conducted at the knee joint the novel analysis tool significantly decreased the standard error of measurement and improved the sensitivity in detecting minor structural changes. It further eliminated the need of time-consuming user interaction, and thereby increasing reproducibility.
影响定量磁共振成像（qMRI）中骨结构分析精度的一个主要因素是在进行纵向研究时勾画和追踪感兴趣区域的个体间和个体内的变异性。本文提出了一种自动分析工具，具有骨髓分割、感兴趣区域生成和关节松质骨特征化的功能。在膝关节评估研究中，该新型分析工具显著降低了测量的标准误差，并提高了检测微小结构变化的灵敏度。它进一步消除了耗时的用户交互的需要，从而提高了可重复性。
