This paper is published in Volume-3, Issue-3, 2017
Area
Medical Image Processing
Author
Rehna Kalam
Co-authors
M. Abdul Rahman
Org/Univ
University of Kerala, Thiruvananthapuram, Kerala, India
Pub. Date
25 May, 2017
Paper ID
V3I3-1357
Publisher
Keywords
Skull stripping, Filtering, Image Registration, Feature Extraction, Classification, Segmentation

Citationsacebook

IEEE
Rehna Kalam, M. Abdul Rahman. Tumor and Edema Segmentation Using Efficient MFCM and MRG Algorithm, International Journal of Advance Research, Ideas and Innovations in Technology, www.IJARIIT.com.

APA
Rehna Kalam, M. Abdul Rahman (2017). Tumor and Edema Segmentation Using Efficient MFCM and MRG Algorithm. International Journal of Advance Research, Ideas and Innovations in Technology, 3(3) www.IJARIIT.com.

MLA
Rehna Kalam, M. Abdul Rahman. "Tumor and Edema Segmentation Using Efficient MFCM and MRG Algorithm." International Journal of Advance Research, Ideas and Innovations in Technology 3.3 (2017). www.IJARIIT.com.

Abstract

Momentarily, categorizing of brain tumor and segmentation is truly an exciting task in MRI. Numerous researches work in generating divergent plus interesting techniques and algorithms for this specified work of medical image processing. On behalf of enhancing a precise brain tumor extraction, we provide an effective methodology for both classification and segmentation i.e., separation of brain MRI images as well as labeling of brain MRI images in terms of edema, tumor, white matter (WM),gray matter (GM) plus cerebrospinal fluid (CSF). At this instant and in our recommended system of brain tumor detection encompasses six segments, i.e. pre-processing, filtering, Image registration, Feature extraction, Classification, Segmentation. At this moment, in case of preprocessing, the input MRI image is firstly fetched from the MRI database and as well subjected to skull stripping for rejecting the undesirable area from the image. In addition, by the utilizing Gaussian filter, the skull stripped image has been smoothened. Subsequently by utilizing Automatic image registration the filtered images are recorded into one coordinate system wherever the movement of the head is a situation often encountered during the imaging process. Shape, intensity and texture are the features that will be extricated from the registered images.On the basis of extricated features, the Brain MRI images are characterized into normal or abnormal images. Finally by utilizing the modified FCM segmentation algorithm tumor portion is extracted and edema is segmented applying modified region enhancing from the abnormal images. Therefore in case of normal image, the Gray matter, the white matter and the cerebrospinal fluid can be segmented. The outcomes are analyzed for illustrating the representation of the suggested classification plus segmentation methodology with prevailing techniques.
Paper PDF