Segmentation and Classification Method in IVOCT Images
Cardiovascular disease (CVD) is a fatal disease of the heart or blood vessels. Intravascular optical coherence tomography (IVOCT) as a newly emerging optical-based technology can provide real-time, high-resolution, and three dimensional images with micrometer resolution. In this paper, an automatic lumen detection method composed of OTSU threshold and active contour model, was investigated to improve the robustness and accuracy. The proposed method is compared with manual lumen detection (MLD), and then average distance and max distance results are obtained. For the given datasets, the average distance and max distance is 0.020mm and 0.088mm respectively. Furthermore, an automatic plaque segmentation and classification is proposed to use Hidden Markov Models(HMM), GLCM and Random Forests algorithm. From the color-code plaque classification results, the approach proposed is available. In conclusion, this method can deal with IVOCT image with high robustness and accuracy.
intravascular optical coherence tomography lumen detection plaque classification
Zhou Ping Zhu Tongjing Li Zhiyong
School of Biological Science&Medical Engineering, Southeast university, Nanjing, 210096, china
国际会议
重庆
英文
327-330
2015-12-18（万方平台首次上网日期，不代表论文的发表时间）