Journal of Northeastern University Natural Science ›› 2020, Vol. 41 ›› Issue (3): 326-331.DOI: 10.12068/j.issn.1005-3026.2020.03.005

• Information & Control • Previous Articles     Next Articles

Knee MRI Segmentation Algorithm Based on Chaotic Moth-Flame Optimization

WANG Hai-fang, QI Chao-fei, ZHANG Yao, ZHU Ya-kun   

  1. School of Control Engineering, Northeastern University at Qinhuangdao, Qinhuangdao 066004, China.
  • Received:2019-08-04 Revised:2019-08-04 Online:2020-03-15 Published:2020-04-10
  • Contact: WANG Hai-fang
  • About author:-
  • Supported by:
    -

Abstract: The moth-flame optimization (MFO) algorithm may show shortcomings such as the local optimum and convergence stagnation when solving the practical optimization problem. Therefore, aiming at the problem that MRI (magnetic resonance imaging) images are difficult to segment, this paper proposes a chaotic moth-flame optimization(CMFO) algorithm. In order to help doctors read the MRI films and improve the efficiency and accuracy of diagnosis, the knee MRI images are selected as research objects during the experiments. Then,CMFO algorithm and maximum threshold entropy are combined and applied into multi-threshold segmentation. In order to present the advantages of the CMFO algorithm proposed, SOA, BFOA and MFO algorithms are introduced under the same condition for comparative experiments. The experimental results show that CMFO can effectively improve the optimal performance of MFO, and has better applicability and advantages for knee MRI image segmentation.

Key words: chaotic strategy, knee MRI image, maximum threshold entropy, multilevel-threshold segmentation, moth-flame optimization (MFO)

CLC Number: