A Novel Segmentation Approach for Brain Tumor in MRI
Yong Wei
1
, H. Keith Brown
2
, Junfeng Qu
3
1.
Department of Computer Science & Information Systems
University of North Georgia, Dahlonega, Georgia, U.S.A.
ywei@ung.edu
2.
Department of Bio-Medical Sciences
Philadelphia College of Osteopathic Medicine-Georgia Campus, Suwanee, Georgia, U.S.A.
keithbr@pcom.edu
3.
Department of Computer Science & Information Technology
Clayton State University, Clayton, Georgia, U.S.A.
junfengqu@clayton.edu
ABSTRACT
Brain MRI image segmentation is one of the most
important applications of image segmentation
technique, and is an important part of clinical
diagnostic tools. Segmented image can help
physicians to identify tumor tissues in brain, and
monitor effectiveness of chemotherapy treatments.
However, manual segmentation of muscle regions is
not only inaccurate, but also time consuming. In this
work, Intensity Space Map (ISM) is used along with
fuzzy c-means clustering algorithm to segment
tumor regions in color MRI images. Experiments
show the proposed ISM-based fuzzy c-means
clustering brain MRI image segmentation yields
promising results.
KEYWORDS
MRI Image; Segmentation; Intensity Space Map
(ISM); Fuzzy C-means Clustering; Brain Tumor.
1 INTRODUCTION
Image segmentation is to group pixels into
regions for future process. In each partitioned
region of an image, pixels have similar
characteristics based on given criteria. Brain
MRI image segmentation is one of the most
important applications of image segmentation
technique, and is an important part of clinical
diagnostic tools. Physicians can use magnetic
resonance images (MRI) to estimate volume of
tumor tissues in brain before and after
chemotherapy. The thickness of reconstruction
of MRI slices is determined during scanning. If
the size of tumor region in MRI can be
measured, the volumetric estimation of tumor
can be obtained by calculating the sum of the
products of the slice thickness and the tumor
region size of each MRI slice. Hence,
segmenting the tumor region in MRI images
becomes the key step in the procedure.
There have been many research endeavors to
segment brain tumor in MRI. Stadlbauer et al.
used the Gaussian distribution of spatial
distribution of choline-containing compounds
(Cho), creatine (Cr) and N-acetyl-aspartate
(NAA) in brain tumors as threshold in normal
brain T2-weighted MRI [1]. The region
growing segmentation methods are part of the
region-based methods, and are the most
commonly used for brain tumor segmentation
[2]. Chong et al. [3] used region growing based
algorithm to measure the tumor volume
demonstrated on pretreatment T2-weighted
magnetic resonance data sets.
The k-means algorithm is the most commonly
used clustering algorithm since it is easy to
implement and found to be effective in many
applications. The fuzzy version of k-means
clustering (fuzzy c-means, FCM) is widely
adopted for medical image segmentation
Proceedings of the The Third International Conference on Digital Enterprise and Information Systems, Shenzhen, China, 2015
ISBN: 978-1 -941968-10-9 ©2015 SDIWC 98