基于置信连接的CT血管造影分层可视化研究

Hierarchical visualization of CT angiography based on confidence-connected

为实现利用CT血管造影技术(CTA)数据构建人体脊柱、血管的分层可视化,充分研究不同组织灰度分布的特征实现典型组织识别,针对组织灰度重叠的关键问题,构建基于置信连接的三维区域生长算法实现重要组织血管和骨骼的剥离,最后用体绘制技术进行三维重建.本文实现了人体腹部皮肤、肌肉、骨骼、血管的分层可视化,各结构可单独显示,可任意旋转,放大缩小.通过调节参数和三维分割技术可以很好地完成分层可视化的目的,可以为医生提供更清晰直观的三维模型.

In order to achieve the hierarchical visualization of the human spine and blood vessels by using the CTA data,the recognition of the typical organization can be realized by properly vtilizing the intensity distribution of different organizations.By tacking the key problem of organization gray level overlapping,the separation of important organizations blood vessels and bone can be realized by three-dimensional region growing algorithm based on confidence-connected.Finally,the three-dimensional reconstruction is performed by volume rendering.The hierarchical visualization of the abdominal skin,muscle,bone and blood vessels is realized,and the structure can be displayed separately,can be freely rotated and zoomed in and out.By adjusting the parameters and three-dimensional segmentation,hierarchical visualization can be easily achieved.So it can provide a more clear and intuitive three-dimensional model for the doctor.