Skip to main content

Research Repository

Advanced Search

A novel model and method based on Nash equilibrium for medical image segmentation

Zhang, Tian; Zhang, Jing; Zhang, Jian; Smith, Melvyn; Hancock, Edwin

Authors

Tian Zhang

Jing Zhang

Jian Zhang

Profile Image

Melvyn Smith Melvyn.Smith@uwe.ac.uk
Research Centre Director Vision Lab/Prof

Edwin Hancock



Abstract

Accurate image segmentation is a very important task in medical image analysis as it can help us to better distinguish tumours from normal tissues. One of the important features of MRI images of glioma (a kind of brain tumour) is that the tumour shapes are most often appear irregular and their contours indistinct. As such, nodes on the contour cannot be easily established and clustered together. In order to cluster a node sets and so segment the glioma image, a novel model together with the method of Nash equilibrium is put forward. Firstly, a model of the Nash equilibrium with double allocation constraints is proposed. Secondly, the principle and formula of the Nash equilibrium based on entropy and standard deviation is proposed. Finally, the determination of the penalty parameter in SVM, using the novel Nash equilibrium to help cluster and segment the glioma image is presented. Experimental results demonstrate that the proposed model and method outperforms other competing methods. It is shown that the method can accurately and correctly segment glioma images.

Citation

Zhang, T., Zhang, J., Zhang, J., Smith, M., & Hancock, E. (2018). A novel model and method based on Nash equilibrium for medical image segmentation. Journal of Medical Imaging and Health Informatics, 8(5), 872-880

Journal Article Type Article
Acceptance Date Oct 11, 2017
Publication Date Jun 1, 2018
Deposit Date Mar 15, 2018
Journal Journal of Medical Imaging and Health Informatics (JMIHI)
Print ISSN 2156-7018
Publisher American Scientific Publishers
Peer Reviewed Peer Reviewed
Volume 8
Issue 5
Pages 872-880
Keywords segmentation, cluster, Nash equilibrium, entropy, SVM
Public URL https://uwe-repository.worktribe.com/output/1435352
Publisher URL http://www.aspbs.com/jmihi/contents_jmihi2018.htm#v8n5