You are on page 1of 5

SEMESTER PROJECT

REPORT

BRAIN TUMOR DETECTION


USING ANN

BY
JAMIL AHMAD
MUHAMMAD ALI
INTRODUCTION
Brain tumor is one of the most dangerous diseases
which require early and precision detection and most of its detection
methods depend on decision of neuro-specialists , and radiologist for
image evaluation which possible to make misdiagnosis due to human
errors and also requires more effort and a long time to detect it. The
main purpose is to design automatic image processing system to detect
brain tumor abnormality using artificial neural networks. This
application will be helpful in hospitals and clinics.

With the advances of digital image processing, radiologists have a


chance to improve their performance with automatic methods like
computer-aided detection (CAD) system and Artificial Neural Networks.
Computer-aided diagnosis (CAD) aims to increase the predictive value
of the technique by pre reading medical images to show the locations
of suspicious abnormalities, and analyze their characteristics, as an aid
to the radiologist.
ABOUT THE ARTICLE
The article was presented by Hussna Elnoor Mohammed Abdalla in
supervision of Dr. Mohamed Yagoub Esmail , this article presents the
idea of detection of tumors in brain by using Artificial Neural Networks
and the input for the neural network is feature of MRI images, these
features will be extracted using Harlick’s equation. In neural network
BPP is used.

BASIC FLOW CHART.


ABOUT OUR APPROACH
As a semester project we had tried to implement the basic form of this
article using the basic functions of matlab. In our approach the system is
detecting the tumor on basis of density (in sense of intensity levels).

Step by Step Implementation Detail

1. Reading an image

2. Converting it to binary with thresholding value 0.7

3. Finding connected components

4. stats= regionprops (BW, properties) measures a set of properties


for each connected component (object) in the binary image, BW.

as properties we need Solidty and Area

'Solidity' — Scalar specifying the proportion of the pixels in


the that are also in the region

'Area' — Scalar; the actual number of pixels in the region.


Then we create a GUI in matlab as shown

You might also like