基于相似性测度和似然模型的血管造影图像分割新方法

Extracting blood vessel from X-ray angiogram based on Bayesian tracking model

由于造影图像中的血管具有复杂的形态结构,经典的基于跟踪的一类方法在进行分割时容易丢失目标,因此其自动化程度和鲁棒性受到了严重的影响。提出了一种鲁棒的最大似然血管跟踪模型,该模型是建立在多种局部相似性测度和智能知识引导的基础上的,能够精确估计出血管轴线,并能正确判断分支点位置。实验结果表明,该模型具有很高的稳定性,能够为血管三维重建和临床诊断提供正确的依据。

The blood vessels in the angiograms take on complicated modalities. The classic tracking-based methods often lose the object in tracking the vessel tree, so that the capabilities of automatization and robustness are greatly affected. A robust Bayesian tracking model is presented to estimate the vessel axis and infer the vessel structure, which is based on multi-comparability＇ measurement and intelligent guidance. The experiment results reflect high stability of the model. The proposed method can provide good reference for 3D reconstruction and clinical diagnosis.