基于Delaunay剖分的心内膜表面动态三维重建算法

Delaunay based dynamic 3D endocardium surface reconstruction algorithm

心内膜表面三维重建技术在三维标测系统中起着手术导航和靶点定位的作用。针对心内科手术中实时采集的散乱心内膜点云,提出了一种基于Delaunay剖分的表面动态三维重建算法。以CGAL非递归方式实现的逐点插入计算Delaunay剖分算法为基础,在剖分过程中,用关联采样点的伞局部替换原来表面中不满足Gabriel准则表面面片的集合,心内膜表面结构随着点云Delaunay剖分的变化而进行动态的更新;同时为了有效地表达心内膜表面及其点云的Delaunay剖分,并能够快速地索引四面体网格和表面三角面片,提出了一种以vtkData Array为基础的几何数据结构。最后,实验表明该方法在重建结果和重建时间上可以满足心内科手术中的临床应用。

When applying surgery navigation and target positioning,the reconstruction of endocardium plays an important role in the 3D mapping system.Given a set of unorganized endocardial points acquired in the surgery,this paper proposed a Delaunay based method to reconstruct the endocardial surface dynamically.The algorithm was based on the method of computing Delaunay triangulation with non-recursive provided by CGAL.During computing Delaunay triangulation,this algorithm used the umbrella incident to a sample point to replace the union of triangles in the surface that were against Gabriel property.Thus,the endocardial surface updated with the variation of Delaunay triangulation.In order to express the topological features of the reconstructed surface and 3D Delaunay triangulation in the memory effectively,it designed a vtkDataArray-based data structure to help index the geometrical cells quickly.Finally,experiments show that this proposed algorithm can meet the clinical application of cardiology both in reconstruction results and reconstruction time.