基于 OCT 图像的冠状动脉血管支架自动检测

摘要

光学相干断层扫描技术（Optical Coherence Tomography ，OCT ）优良的分辨率促使它可以用于指导经皮冠状动脉的介入治疗，包括对支架置入术后即刻的观测和对支架置入后远期血管内修复的评价等。通过使用 OCT 技术对植入支架后的冠脉血管成像，观测者可以评估支架是否很好地与血管壁贴合，以及是否存在支架位置覆盖、支架内再狭窄等问题，这些对于临床诊断和治疗都有重要意义。然而，手动识别所有的这些支架是一项耗时且乏味的工程。基于此，论文通过对植入支架的血管横截面 OCT 图像的特点分析，提取有效的特征，结合随机森林分类器，提出了一套全新的血管支架自动检测方法。它可以很有效地检测出冠状动脉 OCT 图像中支架的位置。%The excellent resolution of Optical coherence tomography(OCT ) can be used to guide interventional therapy of percutaneous coronary artery ,including instant observation after stent implantation and long-term evaluation of stent re-pair etc .According to imaging the coronary angiography after stent implantation though OCT ,the observer can evaluate that whetherthe stentis well fit with the blood vessel wall ,its position covering stent restenosis and so on .Therefore the clinical diagnosis and treatmentsare of great significance .However ,the manual recognition of all these stents is a time-consuming and tedious work .Based on this ,this article analyzes the characteristics of vascular cross-sectional OCT images of stent im-plantation ,extracts its features effectively ,avails the random forest classifier ,and finallyputs forward a new method of auto-matic detection of vascular stent .It can detect coronary OCT images ofthe position of the stent effectively .