Hybrid level set-based method for segmentation of glioma(PDF)
《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]
- Issue:
- 2019年第4期
- Page:
- 414-419
- Research Field:
- 医学影像物理
- Publishing date:
Info
- Title:
- Hybrid level set-based method for segmentation of glioma
- Author(s):
- HE Jianglin; LIU Shiwei; WANG Xingyue; YUE Qing; WANG Yuanjun
- Institute of Medical Image Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China
- Keywords:
- ; Keywords: segmentation of glioma; geodesic active contour model; local image fitting model; level set algorithm; C-V model
- PACS:
- R739.41；R318
- DOI:
- DOI:10.3969/j.issn.1005-202X.2019.04.009
- Abstract:
- Abstract: The boundary information and regional information of glioma images are comprehensively considered in this study, and a segmentation method based on hybrid level set is proposed by combining geodesic active contour model with local image fitting model based on level set. Firstly, the magnetic resonance image of glioma is pre-processed, and the brain tissues are extracted using C-V model. Then a hybrid level set model is established to segment the glioma in the pre-processed image. The experiment proves that the proposed segmentation method can not only simplify the regularization process of the level set function which is initialized to symbol distance function, but also effectively overcome the boundary leakage problem of geodesic active contour model at weak or discrete edges, thereby obtaining better segmentation results.
Last Update: 2019-04-23