IRJET - Clustering Algorithm for Brain Image Segmentation

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 4035
Clustering Algorithm for Brain Image Segmentation
Ashwini Honrao1, Preeti Gore2, Prajakta Kurle3, Priyanka Sairise4, Prof. Ashwini Dhoke5
1,2,3,4,5Dept. of Computer Engineering, Dr. D. Y. Patil Institute of Technology, Maharashtra, Pune
----------------------------------------------------------------------***---------------------------------------------------------------------------
Abstract - Brain tumour is one of the dangerous and critical
disease which is even hard to getdetected withoutany medical
test for example CT Scan, MRI Scan. Image segmentation is a
process that works by segregating any arbitrary image into
non-intersecting regions. The regions obtained after the
division should be such that each region is homogenous and
the union of any two adjacent regions is heterogeneous. Anew
image segmentation is a method by combining the FCM
clustering algorithm with a rough set theory. The rough-fuzzy
c-means algorithm is presented for segmentation of brain
images. The main focus of the work, based on human MRI
brain image, is to optimize the segmentation process with
higher accuracy rate. Cluster analysis recognizescollectionsof
comparable objects and thereforehelps inlearningcirculation
of outlines in big data sets. Clustering is most widely used for
real world applications. The eﬀectiveness of the FCM
algorithm, along with a comparison with other related
algorithms, is demonstrated on a set of brain images. Rough
set theory can be useful method to overcome such
complication during image segmentation. Because of this we
will use FCM clustering algorithm with a rough set theory.
Key Words: MRI Image, Brain Tumour Detection, Image
Processing, Machine Learning, Clustering, Fuzzy C-
means.
1. INTRODUCTION
In computer vision, image segmentation is the process of
partitioning a digital image into multiple segments (sets of
pixels, also known as image objects). The goal of image
segmentation is to simplify or change the representation of
an image into something that is more meaningful and easier
to analyse. Image segmentation is typically used to locate
objects and boundaries (lines, curves, etc.) in images. Image
segmentation is the process of assigning a label to every
pixel in an image such that pixels with the same label share
certain characteristics. The result of imagesegmentationisa
set of segments that collectively cover the entire image, or a
set of outlines extracted from the image.Eachofthepixelsin
a region are similar with respect to some characteristic or
computed property, such as colour, intensity, or texture.
Adjacent regions are significantly different with respect to
the same characteristics.
Picture division is a procedure that works by isolating any
self-assertive picture into non-meeting areas. The areas got
after the division ought to be to such an extent that every
locale is homogenous and the association of any two
contiguous districts is heterogeneous. A new image
segmentation is a method by combining the FCM clustering
calculation with a harshsethypothesis.Theunpleasantfluffy
c-implies calculation is introduced for division of mind
images. The fundamental focal point of the work, in view of
human MRI cerebrum picture, is to improve the division
procedure with higher exactness rate. Bunch examination
perceives assortments of practically identical articles and
thusly helps in learning course of layouts in huge
informational collections. Grouping is most broadly utilized
for true applications.The electiveness oftheFCMcalculation,
alongside an examination with other related calculations, is
shown on a lot of mind pictures. Rough sethypothesiscanbe
helpful technique to beat such intricacy during picture
division. On account of this we will utilize FCM bunching
calculation with an unpleasant set hypothesis.
1.1 Basic Concepts
Machine learning is a branch of Artificial Intelligent
based on the idea that the system can learn from data,
identify a pattern and make a decision with minimum
human interaction.
Image processing is a method to perform some operations
on a image, in order to get an enhanced image or to extract
some useful information from it. It is a type of signal
processing in which input is an image and output may be
image or characteristics or features associated with that
image. FCM means Fuzzy C-means it is an algorithm used for
clustering. Fuzzy C-means is a method of clustering which
allows one piece of data to belong to two or more clusters.
2. LITERATURE SURVEY
This section of Literature Survey eventually reveals some
fact of predication of students performance based on the
analysis of many authors work as follows:
1. In this paper A. Namburu, S. K. Samay, and S. R. Edara [1]
proposed a FCM image segmentation algorithm is to divide
the sample points of vector space intoCsubspacesaccording
to some distance measure. Segmenting the MRI image
particularly with weak object boundaries, which is very
difficult for fuzzy c-means and level set method to apply
image segmentation
2. X. Yong, Z. Ji, and Y. Zhang [2] the author’s goal is to Fuzzy
C-means Clustering (FCM) algorithm is one of the most
classical fuzzy clustering algorithms, which searches for the
optimal extremes through repeated iterations. Some
partitions utilized in the segmentation have an enormous
impact on the output.
3. Herng-Hua Chang and Chih-Chung Hsieh [3] explains
comparisons with two famous methodsand evaluatedonthe

internet brain segmentation repository (IBSR) datasets.
Fully automatic methodssometimes fail,producingincorrect
results and requiring the intervention of a human operator.
This is often true due to restrictions imposed by image
acquisition, pathology and biological variation.
4.Rosita Shishegar1,2 Anand A. Joshi3 Mary Tolcos4,5David
W. Walker5 Leigh A. Johnston1,6 [4] paper presents a new
atlas-free method that segments the inner and outer
boundaries of the cortical plate in fatal brains by exploiting
diffusion-weighted imaging cues and using a cortical
thickness constraint. Investigated the potential of deep
architecture through small kernel by comparing deep CNN
shallow architecture with larger filters that shallow
architecture presented a lower performances.
5.Christoph Matthies, Franziska Dobrigkeit, Guenter Hesse
[5] defines a parameter-efficient and practical fully
convolutional neural network architecture, termed InfiNet,
aimed at voxel-wise semantic segmentation of infant brain
MRI images at iso-intense stage, which can be easily
extended for other segmentation tasks involving multi-
modalities. One of the main problems in medical image
segmentation is uncertainty. Some of the sources of this
uncertainty include
6. Lata Ayesha Akter and Goo-Rak Kwon [6] This paper
proposed a new integrated image segmentation method for
MRI brain images. In this method we have used a new
transformation called Contourlet Transform which is
integrated with canny edge detector. Using traditional
algorithm if the boundaries of the image are blur then it
becomes difficult for segmentation
7. K. Y. Lim and R. Mandava [7] The results of the fuzzy
clustering are de-fuzzified, and the fuzzy clustering is
transformed into deterministic classification. Problem has
been to combine two or more segmentation methods in
order to integrate and merge their advantages for the
purpose of dealing with the different characteristics and
variability of the data.
2.1 Relevant Mathematical Model
A binary masking image is establishedtodetectlargetargets,
which is given by formula (1). The background image can be
obtained by the product of the masked image and thesource
image, which can be calculated by formula (2)
Bm = I >Bth…..(1)
IB=Bm×I…..….(2)
Where Bm is a binary masked image, I is MR image,
B th is the threshold of image binarization. It is obtained by
Otsu threshold method IB is a black background image.
Where,
'n' is the number of data points.
’vj’ represents the jth cluster center.
'm' is the fuzziness index m € [1,∞].
'c' represents the number of cluster center.
'µij' represents the membership of ith data to jth cluster
center.
'dij' represents the Euclidean distance between ith data and
jth cluster center.
A finite set X ={x1,x2,···xn}is a set of n samples and c is a
predetermined number of classes; mi(I=1,2···c) is the center
of each cluster µj(xi) is the membership degree of the first
sample with respect to class j. The clustering criterion
function is defined as a membership function:
Where ||xi-mj||2 is the Euclidean distance. b is a fuzzy
weighted power index, which can control the degree of
ambiguity of clustering results. M is the fuzzy c partition
matrix of X. V is the cluster center set of X. The result of
clustering algorithm is to obtainMandVwhichminimizethe
criterion function. In the method of fuzzy mean clustering,
the sum of membership degrees of samples to eachclusteris
required to be 1:
Let the partial derivative of J(M,V) to mj and µj(xi) be zero.
The necessary condition for obtaining the minimum is:

3. SYSTEM ARCHITECTURE
Fig -1: System Architecture
3. ADVANTAGES:
[1] The method is simple and has advantages in processing
speed, and it is suitable for the segmentation of blurred
boundary areas in images.
[2] The main advantage of automating the process is that it
has made the processing much less tedious and not at all
time consuming.
[3] To developed will try to give best Segmentation and the
detection of disease. It is having more accuracy than K-
Means algorithm
[4] Gives best result for overlapped dataset and
comparatively better than K-Means algorithm.
[5] Unlike K-Means where data point must exclusively
belong to one cluster centre here data point is assigned
membership to each cluster centre as a result of which data
point may belong to more than one cluster centre.
4. CONCLUSION
Fuzzy C-means found great in terms of accuracy and work
efficiency. Image Segmentation will improve the life style
and livingness of human. It will help in medical relatedfields
and also handle major work load of human power. Image
processing is one of them. Data Pre-processing is one of the
most important part of Disease Detection fromimage.Image
Segmentation is the process of partitioning a digital image
into multiple segmentation. Using fuzzy c means algorithm
image segmentation become simple, fast and generate more
accurate result.
REFERENCES
[1] A. Namburu, S. K. Samay, and S. R. Edara, “Soft fuzzy
rough set-based MR brain image segmentation”, May-2017.
[2] X. Yong, Z. Ji, and Y. Zhang, “Brain MRI image
segmentation based on learning local variational Gaussian
mixture models”, Sep. 2016.
[3] Herng-Hua Chang and Chih-Chung Hsieh, “Brain
segmentation in MR images using a texture-based classifier
associated with mathematical morphology”,2017.
[4] Rosita Shishegar1,2 Anand A. Joshi3 Mary Tolcos4,5
David W. Walker5 Leigh A. Johnston1,6, “Automatic
segmentation offetal brainusingdiffusion-weightedimaging
cues”,2017.
[5] Christoph Matthies, Franziska Dobrigkeit, Guenter
Hesse,“InfiNet Fully convolutional networks for infant brain
MRI segmentation”,2018.
[6] Lata Ayesha Akter and Goo-Rak Kwon, “Integration of
Contourlet Transform and Canny Edge Detector for Brain
Image Segmentation”,2018.
[7] K. Y. Lim and R. Mandava, “A multi-phasesemi-automatic
approach for multisequence brain tumor image
segmentation”,Dec 2018.

IRJET - Clustering Algorithm for Brain Image Segmentation

Related slideshows

More Related Content

IRJET - Clustering Algorithm for Brain Image Segmentation