Report templet 2

overratedbeltΤεχνίτη Νοημοσύνη και Ρομποτική

25 Νοε 2013 (πριν από 3 χρόνια και 11 μήνες)

119 εμφανίσεις



SSGMC
E, Shegaon


1

Chapter 1

Introduction


The brain is the anterior most part of the central nervous system. Along with
the spinal cord, it forms the Central Nervous System (CNS). The Cranium, a bony
box in the skull protects it. Virtually everything we do, think, act, rea
son, walk, talk,
the list is endless is because of our brain. Succinctly put, the brain is our survival kit.

Brain Tumors are one of the diseases caused in
the brain. Brain tumors are the
tumors
that grow in the brain. Tumor is an abnormal growth caused by

cells
reproducing themselves in an uncontrolled manner. A
benign brain tumor

consists
of benign (harmless) cells and has distinct boundaries. Surgery alone may cure this
type of tumor. A
malignant brain tumor

is life
-
threatening. It may be malignant
becau
se it consists of cancer cells, or it may be called malignant because of its
location. A malignant brain tumor made up of cancerous cells may
spread or seed

to
other locations in the brain or spinal cord. It can invade and destroy healthy tissue so
it cann
ot function properly.

The structure and function of the brain can be studied noninvasively by
doctors and researchers using Magnetic Resonance Imaging (MRI). The MRI image
in Figure 1.1 is actually a thin horizontal slice of the brain.

The white area at l
ower
left is the tumor. It looks white because MRI scans enhance tissue differences. The
tumor is actually on the right side of the brain. It appears to be on the left side here
because the MRI view here is from below, looking up.



Figure 1.1: MRI Brain

Tumor Image



SSGMC
E, Shegaon


2


Magnetic Resonance Imaging (MRI), strongly depend on computer technology to
generate or display digital images. Segmentation is an important process in most
medical image analysis. Clustering to magnetic resonance (MR) brain tumors
main
tains efficiency. Clustering is suitable for biomedical image segmentation as the
number of clusters is usually known for images of particular regions of the human
anatomy.


In this project we use luminosity
-
based segmentation method. This
project

analyses various clustering

techniques to locate

tumor objects in Magnetic Resonance
(MR) brain images. The input is the MRI image of the axial view of the human brain.
The Clustering algorithms used are K
-
means Clustering, Self Organizing Map (SOM)

and F
uzzy C
-
Means Clustering

(FCM)
. The contrast of the
given gray
-
level MR
image
is adjusted
and
then
clustering algorithms are applied. The position of tumor
objects is separated from other items of an MR image by using clustering algorithms
and histogram
-
clu
stering. This method distinguishes exactly the lesion.

In this system, we combine,




Various Clustering Algorithms one by one and




Histogram Clustering



After the clustering process, the cluster containing the tumor is selected as the
primary seg
ment. To eliminate the pixels which are not related to the tumor pixels,
Histogram clustering is applied.


The performance analysis is conducted by taking a MRI Brain tumor image as
the

input and applying all the three
clustering algorithms to th
e image. The
performance of the above four clusterin
g algorithms are found based on t
he number of
tumor pixels
. The efficiency of all the three

algorithms in this system is found by
applying all the four algorithms to a database of
20

images.


The r
est of the document is organized as follows. Chapter 2 explains the litera
ture
review, Chapter 3 describes the medical imaging techniques
, Chapter 4 gives the
segmentation methods
, Chapter 5 gives the
clustering algorithms

step by step and
their
implem
enta
tion using MATLAB, Chapter 6

gives the
experimental
results and
finally Chapter
7

discusses the conclusion and future
scope
.





SSGMC
E, Shegaon


3

1.1


Problem Definition

There are different diagnosis methods are available among that biopsy is the only sure
way to diagnose a br
ain tumor. In biopsy surgeons removes the tissue for tumor cells.
A pathologist looks at the cells under a microscope to check for abnormal cells
. A
biopsy can show cancer, tissue changes that may lead to cancer, and other
pathological conditions. So befor
e biopsy we must have the proper location of the
abnormal tissues. Also after biopsy
when doctors go for surgery, they must know the
tumor extent in the brain. To determine this we have to use the image processing of
MRI brain images. Finally the problem d
efinition is given below:





It is very difficult to conduct surgery without using image processing techniques.



Structures like tumor, brain tissue and skull cannot be identified without image
segmentation.



In MRI images, the amount of data is too much for

manual interpretation and
analysis.



It takes a long time for diagnosis without image processing techniques.



There is a chance of wrong diagnosis without image processing techniques.


1.2


Objective of the Project

The objective of this project is given below:



To locate the brain tumor in axial view of brain MR images using the luminosity
-
based segmentation.



To segment of MR images using unsupervised clustering methods:

o

K
-

means clustering

o

Fuzzy C
-

Means clustering

o

Self Organizing Maps (SOM) clustering



To find
out the performance analysis of these three clustering methods on the
basis of number of pixels in segmented cluster of brain tumor

and the execution
time required for the same
.










SSGMC
E, Shegaon


4


…….other chapters


































SSGMC
E, Shegaon


5

References


1.

Y.Murachi

etal. “A 95mW MPEG2 MP@HL Motion Estimation Processor
Core for Portable High Resolution Video Application”,
IEICE Transactions
on Fundamentals of

Electronics, Communications and Computer Sciences
archive

Volume E88
-
A , Issue 12, December 2005

2.

Li, W. Li, W
.; Salari, E., "Successive elimination algorithm for motion
estimation,

"
Image Processing, IEEE Transactions on
, vol.4, no.1, pp.105
-
107, Jan 1995

3.

Kim J.S.; Park, R.
-
H., "A fast feature
-
based block matching algorithm using

integral projections,"
Selected
Areas in Communications, IEEE Journal on
,
vol.10, no.5,

pp.968
-
971, Jun 1992

4.

Jong
-
Nam Kim; Tae
-
Sun Choi, "A fast three
-
step search algorithm with
minimum

checking points using unimodal error surface assumption,"
IEEE

Transactions on Consumer Electronics,

, vol.
44, no.3, pp.638
-
648, Aug 1998.

5.

Lai
-
Man Po; Wing
-
Chung Ma, "A novel four
-
step search algorithm for fast
block motion estimation ,"
IEEE Transactions on Circuits and Systems for
Video Technology,

v
ol.6, no.3, pp.313
-
317, Jun 1996

6.

“Technical Note : TN
-
47
-
04 Calculating Memory System Power for DDR2”,

http://www.micron.com
.

7.

“MT46H8M32LF/LG


256Mb, 32 bit Mobile DDR SDRAM”,

http://www.micron.com



Books
:

1.

Fundamentals of digital l
ogic with VHDL design
, Stephen Brown,
McGraw
Hill Publication, 2005

2.

Advanced Signal Processing Handbook
, Editor: Stergios Stergiopoulos,
CRC
Press LLC, 2001


Magazines
:

1.

Electronics for you, Dec 2011