基于软组织形变的高维数据聚类算法研究

Research on High Dimensional Data Clustering Algorithm Based on Soft Tissue Deformation

将高维数据聚类法中子空间凝聚算法用于软组织形变仿真,保证虚拟手术中软组织形变真实性与实时性的均衡。该算法首先进行软组织表面形变区域的映射,而后完成平面圆形区域的划分,最后在平面圆形区域中形成全空间与子空间之间的函数关系。通过Open GL库联合C++语言进行模拟仿真,实验结果表明,该算法简单高效、计算量小、实时性好、真实性高,基本能满足一般软组织浅层形变仿真要求。

The subspace cohesion algorithm in high-dimensional data clustering method is applied to the simulation of soft tissue deformation, which ensures the balance between the authenticity and timeliness of soft tissue deformation in virtual surgery. Firstly, the algorithm maps the soft tissue surface deformation region, then completes the division of the planar circular region, and finally forms a functional relationship between the full space and the subspace in the planar circular region. The simulation results made through the analog simulation with OpenGL and C++ language show that the deformation method proposed in this paper is simple and efficient, with small computation, good timeliness and high authenticity, which can meet the requirements of general soft tissue shallow deformation simulation.