Brain Tumor Segmentation in 3D MRIs Using an Improved Markov Random Field Model
Markov Random Field (MRF) models have been recently suggested for MRI brain segmentation by a large number of researchers. By employing Markovianity, which represents the local property, MRF models are able to solve a global optimization problem locally. But they still have a heavy computation burden, especially when they use stochastic relaxation schemes such as Simulated Annealing (SA). In this paper, a new 3D-MRF model is put forward to raise the speed of the convergence. Although, search procedure of SA is fairly localized and prevents from exploring the same diversity of solutions, it suffers from several limitations. In comparison, Genetic Algorithm (GA) has a good capability of global researching but it is weak in hill climbing. Our proposed algorithm combines SA and an improved GA (IGA) to optimize the solution which speeds up the computation time. What is more, this proposed algorithm outperforms the traditional 2D-MRF in quality of the solution.
Brain tumor Improved Genetic Algorithm Magnetic Resonance Image Markov Random Field Model Simulated Anealing
Sahar Yousefi Reza Azmi Morteza Zahedi
Department of Computer and IT Shahrood University of Technology Shahrood, Iran Department of Computer Alzahra University Tehran, Iran
国际会议
2010 International Conference on Signal and Information Processing(2010年IEEE信号与信息处理国际会议 ICSIP2010)
长沙
英文
382-386
2010-12-14（万方平台首次上网日期，不代表论文的发表时间）