PowerPoint Presentation
A Novel Approach for the Detection and Analysis of Brain Tumour
Under the guidance of Dr. Monica Jain Prof. & HOD-ICE Dept.
GCET Gr. Noida
Submitted By: Mayank Yadav (1109722024) Shivanky Jaiswal
(1109722050) Sandeep Maurya (1109722045)
MD.
Brain Anatomy
What is a brain tumor?A group (mass) of abnormal cells that get
formed in the brain.There are over 120 different types of brain
tumors, which makes effective treatment complicated. Mostly tumors
are classified into two types: (1)Malignant(cancerous) tumors
(2)Benign tumors3Diagnosis of brain tumorsDiagnostic tools include:
patient history, a brain scan, CT scan, MRI. MRI provides a much
greater contrast between the different soft tissues of the body
than computed tomography (CT) does.
4Each year more than 200000 people in world are diagnosed with
primary or metastatic tumour.Brain cancer remains one of the most
incurable forms of cancer, with an average survival period of one
to two years.Technology is useless if it cant even save the
lives.The project aims at early detection of the tumour, better
readability and understandability of MRI reports by each individual
on its own.Automation of tumor detection is required because there
might be a shortage of skilled radiologists at a time of great
need.
MotivationBut what forces us to pursue this project?????
Being an instrumentation engineer, accuracy and precision are
our ethics.
Through this project we aim to do the computational study of MRI
images in a precise and accurate manner & to develop a better
control and detection of the tumour and its cure at early stages.
OBJECTIVETo detect the tumour through edge detection.Design the 2D
structure of tumour to calculate the area.Pre-processing operations
followed by processing operationsCalculation of area of tumor
present in the skull region
MethodologyPreprocessing code
I= imread('Image11.bmp'); % calling the function imread to read
the imageimshow(I); % imshow function called to display the imageJ
= imnoise(I,'salt & pepper',0.02); % imnoise function called to
image with noisefigure, imshow(J) % imshow displays the image with
noiseL = medfilt2(J,[3 3]); % medfilt2 function performs 2-d
filtering on imagefigure, imshow(L)[junk threshold] = edge(I,
'canny'); % edge detection by canny methodfudgeFactor = .5; %
scalar value to be multiplied by threshold valueBWs =
edge(I,'canny', threshold * fudgeFactor); % bws stores the edge
detected imagefigure,imshow(BWs), title('binary gradient mask');%
imshow displays edge detected imagese90 = strel('line', 3, 90); %
creates a structuring element for dilationse0 = strel('line', 3,
0);BWsdil = imdilate(BWs, [se90 se0]); % imdilate function called
to perform image dilationfigure, imshow(BWsdil), title('dilated
gradient mask');BWdfill = imfill(BWsdil, 'holes'); % imfill
performs flood fill operations on background pixels of input
image
Processing codeA=
imread('image11.bmp');B=Imread('image7.bmp');I=B;[T,EM]=graythresh(I);
% EM effectiveness matrix,t normalized intensity value: both
arguments of graythreshg = im2bw(I,2*T); % g stores the thresholded
imagebg_wT=bwareaopen(g,2*500); % bwareaopen function called to
removes from binary image specified connected
componentsbg_withoutT=bwareaopen(bg_wT,2*1000); % variable stores
the image without tumortumor=imsubtract(bg_wT,bg_withoutT); %
imsubtract called to subtract one image from
anothertumor=imfill(tumor,'holes');r=regionprops(tumor); % measures
set of properties associated with imagecent=cat(1,r.Centroid); %
cat argument concatenates the objectsbb=cat(1,r.BoundingBox); %
bounding box creates a box around the specified
imageArea=cat(1,r.Area) % area argument calculates the area of
image in bounding boximshow(I)hold onfor k=1:size(bb,1) % loop for
creating and defining the bounding box
rectangle('position',bb(k,:),'edgecolor','r','linewidth',2)
plot(cent(k,1),cent(k,2),'g+')endfigure,imshow(tumor);bw=imfill(bw,'holes');Main_Area=bwarea(bw);vpa(Main_Area,10)
% variable precision accuracy
Results Input image 1 and its outputs
Input image 2
Input image 3
Input image 4
Calculations Input 1
Total skull area:33635.125Tumor area:1933Percentage of area
occupied by tumor:5.7%
Input image 2
Total skull area:35711.125Tumor area:1716Percentage of area
occupied by tumor:4.8%
Input image 3
Total skull area:32269.875Tumor area:1095Percentage of area
occupied by tumor:3.39%
Input image 4
Total skull area:24306Tumor area:2053Percentage of area occupied
by tumor:8.4%
Conclusion MRI is a critical part in many researches.In this
work, the brain image testing process has been done. This method
has given the reliable result. If the brain image has the tumor
region, the further processing steps are needed to be done. The
edge detection operator technique will detect the size, shape and
boundary extraction. The proposed work provides a new algorithm for
brain tumor which is more efficient. The accurate and precise
detection is crucial otherwise the wrong identification of disease
can lead to several consequences. As diagnosis of tumor is a
complicated task; therefore accuracy and reliability are always
assigned much importance. As shown earlier, various MRI results
have been taken and tested, which shows the clear understanding of
accuracy and precise result.
Thank you!!!!!!
