A Nonlinear Adaptive Level Set for Intravascular Ultrasound Images Segmentation

Abstract

In this paper, a level set method (LSM) with the aim of segmenting lumen and non-lumen pixels and Hidden Markov Random Field (HMRF) with the purpose of computing boundaries of lumen are proposed. This proposed methods was evaluated on IVUS images of 7 patients and also our results have shown that using LSM-HMRF methods leads to increasing accuracy up to 85%. Results also showed that combination of LSM-HMRF could successfully identify the lumen boundary. The main advantage of this method is that one pattern using LSM from all of IVUS images is obtained. The simulation results depicted the effectiveness or the proposed method.