A hybrid approach for detection of brain tumor in MRI images

Author

Abbasi, Solmaz ; Tajeri Pour, Farshad

Author_Institution

Sch. of Electr. & Comput. Eng., Shiraz Univ., Shiraz, Iran

fYear

2014

Firstpage

269

Lastpage

274

Abstract

In this paper, a method for 3D medical image segmentation is presented. This method is used to detect brain tumor in MRI images by combining Clustering and Classification methods to decrease the complexity of time and memory. In the first phase, non-negative matrix factorization with sparseness constraint method is used to separate the region of interest from the image. In the second phase, the classification of the region of interest is performed. For this purpose, TOP-LBP features and gray level co-occurrence matrix are extracted and Random forest is used for classification and segmentation of the necrosis, edema, non-enhanced tumor and enhanced tumor. This method has achieved a fast speed for segmentation of MRI 3D images and has been evaluated with criteria of Dice´s and Jacquard´s coefficient on the brain tumor from magnetic resonance image obtained from the Brats2013 database.

Keywords

biomedical MRI; brain; computational complexity; decision trees; feature extraction; image classification; image segmentation; image texture; matrix decomposition; medical image processing; neurophysiology; object detection; pattern clustering; random processes; statistical analysis; tumours; 3D MRI image; 3D medical image segmentation; Brats2013 database; Dice coefficient; Jacquard coefficient; Random forest; TOP-LBP feature extraction; clustering; edema classification; edema segmentation; fast segmentation speed; gray level co-occurrence matrix extraction; hybrid brain tumor detection; magnetic resonance image; memory complexity reduction; necrosis classification; necrosis segmentation; nonenhanced tumor classification; nonenhanced tumor segmentation; nonnegative matrix factorization; region of interest classification; region of interest separation; sparseness constraint; time complexity reduction; Biomedical engineering; Educational institutions; Feature extraction; Image segmentation; Magnetic resonance imaging; Three-dimensional displays; Tumors; Brain tumor detection and segmentation; magnetic resonance images(MRI); random forest; texture features;

fLanguage

English

Publisher

ieee

Conference_Titel

Biomedical Engineering (ICBME), 2014 21th Iranian Conference on

Print_ISBN

978-1-4799-7417-7

Type

conf

DOI

10.1109/ICBME.2014.7043934

Filename

7043934