An Automatic Stent Detection for Intravascular Optical Coherence Tomographic

Intravascular Optical Coherence Tomography (IVOCT) is an important imaging technique for diagnose of atherosclerosis due to its advantages of high spatial and temporal resolutions. In this paper, an automatic detection method is proposed for the evaluation of the stent placement in Percutaneous Coronary Intervention (PCI). The method mainly contains the flowing three steps. Firstly, the IVOCT images are de-noised by total variation model. Secondly, the ROI zone and the seed points of stents are detected. Finally, the region growing algorithm is used to segment the stents according to the selected seed points. More than one thousand clinical OCT images are collected from 10 post-PCI treatment patients for evaluating the performance of the proposed method. The results show that the proposed method could achieve 89.3% accuracy and 86.5% sensitivity.