Scribd
Upload
4 views

A6 Batch

Uploaded by

Praveen Kumar karu
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
4 views

A6 Batch

Uploaded by

Praveen Kumar karu
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 24

AUTOMATIC SEGMENTATION OF MR

BRAIN IMAGE FOR TUMOUR


EXTRACTION AND EVALUATION

Guided By:
 D . Poojitha Koteswari-19HT1A0421 D . Prabhakar Rao
 K . Praveen Kumar -19HT1A038 Asst professor
 N . Suresh -19HT1A0456
 O. Soma Sekhar -19HT1A0459
CONTENTS
 Abstract
 Introduction
 Objective
 Problem statement
 Tumour detection methods
 Existing system
 Draw backs of existing system
 Proposed System
 Output
 Software and Hardware requirements
 Advantages
 Applications
 Conclusion
 References
ABSTRACT
Automatic segmentation of brain tumors from medical images is important for clinical
assessment and treatment planning of brain tumors.
The brain tumor is a disease that affects or harms the brain with unwanted tissues.
This is very difficult to detect brain tumor tissue from whole brain.
To detect the location and size of brain tumors required MRI images of brain tumors.
INTRODUCTION
MRI images can help differentiate brain tissue, brain tumors, edema,
and cerebrospinal fluid based on differences in color contrast in each
tissue.
Detection or segmentation techniques are used to detect and segment
the brain-tumor region from the MRI images of brain.
The brain-tumor segmentation process is performed for separating
brain-tumor tissues from brain MRI images.
OBJECTIVE
The main objective of the segmentation of tumour extraction and evaluation is to
detect the earlier stage of tumour and provide proper treatment to cure it.
PROBLEM STATEMENT
Tumor detection and removal is one medical issue that still remains challenging in the
field of biomedicine.

 Early imaging techniques such as pneumoencephalography and cerebral angiography


had the drawback of being invasive and hence the CT and MRI imaging techniques
help the surgeons in providing a better vision.
BRAIN TUMOUR CAUSES
Clotting of blood cells.
Improper sleeping hours.
Restless work.

SYMPTOMS OF BRAIN TUMOUR


Heavy headache.
 Vomiting’s.
Rotate (or) memorising of eyes.
TUMOUR DETECTION METHODS
OSTU(Orthogonal segmentation threshold unit)method
Segmentation method
Clustering method
EXISTING SYSTEM
Brain tumour image

Pre-processing

Sharpening

Threshold Segmentation Morphological


Operation

Output image with tumour


detection
DRAWBACKS OF EXISTING SYSTEM
Tumour detection is difficult.
Brightness problem.
PROPOSED SYSTEM
Clustering methods are two types
K-mean clustering method
Fuzzy c-mean method

Segmentation using k-
Input image Pre-processing
mean

Approximate Feature
Clusters
tumour detection extraction
K-MEAN METHOD
Image data

Get k value and choose


random centre

Calculate the mean for each


cluster

Calculate distance between


each pixel and cluster
No Compare to next
If pixel close to cluster cluster

Move to same cluster Repeat process

Re estimate the cluster Clustered image


CLUSTERING PROCESS
FINAL CLUSTERS
MATHEMATICAL EXPRESSION ALGORITHM:
Give the number of cluster value ask.
Choose the k cluster random centre.
Calculate centre and mean of cluster.
Calculate the distance between each pixel to each cluster
centre.
If the distance is near to the centre then move to that cluster.
Otherwise move to next cluster.
Re-estimate the centre.
Repeat the process until the centre doesn't move.
FUZZY C-MEAN METHOD
Image data

Get k value and choose centre

Calculate the mean for each


other cluster

Calculate distance between each


pixel and cluster

No Compare to next cluster


If pixel close to cluster
Repeat process
Move to same cluster
Clustered image
MATHEMATICAL
EXPRESSION ALGORITHM:
Give the number of cluster value ask.
Choose the k cluster random centre.
Calculate centre and mean of cluster.
Calculate the distance between each pixel to each cluster
centre.
Where m=any real number greater than 1
If the distance is near to the centre then move to that cluster.
u ij=degree of membership of x i in the
cluster. Otherwise move to next cluster.
x j=data measured in d- dimensional. Re-estimate the centre.
R j=d-dimension centre of the cluster Repeat the process until the centre doesn't move.
DETECTION OF TUMOUR AREA
Size _ of _tumour , S =[(sqrt of p)*0.264]mm^2
where p=number of white pixels
w=width
h=height 255 255
Number _ of_ white pixel , p= f(0)
W=0 H=0

Where p =number of white pixels(width*height)


1pixel=o.264mm
SOFTWARE AND HARDWARE
REQUIREMENT
Operating system - WINDOWS XP/7
Coding language - MATLAB
Tool - MATLAB R 2009
ADVANTAGES
K-means clustering has the advantage of being automatic, faster in execution and
lesser computational complexity.
Morphological operators based segmentation is more accurate than k-means
clustering, effectively eliminates the noise and in homogeneities due to irregularities in
MR scanner Region growing method, though semi-automatic, gives most accurate
results among three methods.
APPLICATIONS
To detect tumour in brain even if it is in small size.
MRI can help doctors look for conditions such as bleeding , swelling , problems with
the way the brain developed , tumours , infections , inflammation , damage from an
injury or a stroke , or problems with the blood vessels.
CONCLUSION
This paper mainly focuses on identification of brain tumor in an efficient manner with
lesser rate of similarity by using Fuzzy Inference System.
A surgeon can have a clear knowledge about the tumor and its location in the brain.
So the algorithm proposed using Fuzzy Inference System for tumor extraction in MR
brain image has reduced the complexity of the problem and helps the surgeon in locating
the tumor.
The clear anatomical structure of the brain tumor is identified using the proposed fuzzy
based algorithm.
REFERENCES
Boucher A, Pastore J and Ballarin V. Segmentation of medical images using fuzzy
mathematical morphology, JCS and T 2007; 7(3): 256-62.
Wei Sun and Yeoman Wang. Segmentation method of MRI using fuzzy Gaussian basis
neural network neural information processing, Letters and Reviews 2005; 8(2): 19-24.
 Vanier MW, Speidel CM, Rickman DL. Magnetic resonance imaging multi spectral tissue
classification, Journal of NIPS 1991; 10(3): 395-407.
 Pham D.L. Unsupervised Tissue Classification in Medical Images using Edge-Adaptive
Clustering. Proceedings of the 25th Annual International Conference of the IEEE EMBS.
Cancun. Mexico. 2003.

You might also like