Efficient curve-skeleton computation for the analysis of biomedical 3d images - biomed 2010.
Efficient curve-skeleton computation for the analysis of biomedical 3d images - biomed 2010.
复制标题
用于分析生物医学 3D 图像的高效曲线骨架计算 - biomed 2010。
DOI：
--
复制
发表时间：
2010
期刊：
影响因子：
--
通讯作者：
D. Dreossi
中科院分区：
文献类型：
--
作者：
F. Brun;D. Dreossi
磁共振 (MR) 和计算机断层扫描 (CT) 等三维 (3D) 生物医学成像技术的进步使得重建人体部分和其他生物样本的高质量 3D 模型变得容易。主要挑战在于对所得模型进行定量分析，从而能够更全面地表征所研究的对象。一种有趣的方法是基于曲线骨架（或中轴）提取，它提供了有关拓扑和几何的基本信息。曲线骨架已应用于血管网络分析和气管狭窄诊断以及虚拟内窥镜检查中的 3D 飞行路径。然而，曲线骨架计算是一项至关重要的任务。 N. Cornea 在[1]中提出了一种有效的骨架化算法，但它缺乏计算性能。由于成像技术的进步，3D 图像的分辨率越来越高，因此需要有效的算法来分析重要的感兴趣体积 (VOI)。在本文中，提出了一种基于[1]中提出的思想的改进骨架化算法。还报告了原始方法和所提出的方法之间的计算比较。获得的结果表明，所提出的方法允许显着的计算改进，使得在生物医学图像分析应用中采用骨架表示更具吸引力。
Advances in three dimensional (3D) biomedical imaging techniques, such as magnetic resonance (MR) and computed tomography (CT), make it easy to reconstruct high quality 3D models of portions of human body and other biological specimens. A major challenge lies in the quantitative analysis of the resulting models thus allowing a more comprehensive characterization of the object under investigation. An interesting approach is based on curve-skeleton (or medial axis) extraction, which gives basic information concerning the topology and the geometry. Curve-skeletons have been applied in the analysis of vascular networks and the diagnosis of tracheal stenoses as well as a 3D flight path in virtual endoscopy. However curve-skeleton computation is a crucial task. An effective skeletonization algorithm was introduced by N. Cornea in [1] but it lacks in computational performances. Thanks to the advances in imaging techniques the resolution of 3D images is increasing more and more, therefore there is the need for efficient algorithms in order to analyze significant Volumes of Interest (VOIs). In the present paper an improved skeletonization algorithm based on the idea proposed in [1] is presented. A computational comparison between the original and the proposed method is also reported. The obtained results show that the proposed method allows a significant computational improvement making more appealing the adoption of the skeleton representation in biomedical image analysis applications.