Texture Feature Extraction from a mammography Image

I need to extract Texture Features of ROI to be classified whether as benign(normal) or malignant(abnormal).

I used a code, but it's results is not good using different well-known classifiers (SVM kernels for example)

I uploaded a picture that illustrates the following:

1- The first picture titled as('Removed Pectoral'): Breast Image after removing (pectoral muscle)

2- The second picture titled as('ROI'):The Extracted ROI from the removed pectoral muscle image

3- The third picture titled as('segmented ROI): Is the Segmented ROI (Segmentation of ROI image)

My main aim is to obtain the features I uploaded (Additional features in the code)

I used this code for obtaining the features

    % Use the Extracted ROI named as "ROI" in the uploaded Images
    % Use the following angles "offset":(0,45,90,135)
    GLCM = graycomatrix(ROI,'Offset',[0 1;-1 1;-1 0;-1 -1]);

    % Calculate the four built-in MATLAB features
    stats=graycoprops(GLCM,{'Contrast','Homogeneity','Correlation','Energy'});

    contrast=(stats.Contrast);  
    en=(stats.Energy);
    co=(stats.Correlation);
    hom=(stats.Homogeneity);

    % Calculate mean and standard deviation

    m=mean(mean(ROI));    
    s=std2((ROI));

    % The first feature vector 
    f1=[m s hom co en contrast];

    % Calculate the texture Features 
    % Picture used is" segmented ROI" in the uploaded images
    % This picture is the segmentation of " ROI " picture used at the 
    % beginning of the code

    % Features to be extracted " Area, Perimeter, compactness, smoothness, entropy,..."

    J2=SegmentedROI;
    I=J2;  
    J2 = uint8(255 * mat2gray(J2));

    % Detecting the edges of the ROI

    J3=edge(J2,'log');    
    bw=bwareaopen(J2,150);                          
    bwp=edge(bw,'sobel');

    geometric.area=sum(sum(bw));
    geometric.peri=sum(sum(bwp));
    geometric.compact=geometric.peri^2/geometric.area;

    count=1;
    roih=0;
    for i=1:size(bw,1);
          for j=1:size(bw,2);
              if bw(i,j)
                  roih(count,1)=double(SegI(i,j)); 
                  count=count+1;
              end
          end
      end

    nh=roih/geometric.area;
    texture.mean=mean(roih);
    texture.globalmean=texture.mean/mean(mean(double(I)));
    texture.std=std(roih);
    texture.smoothness=1/(1+texture.std^2);
    texture.uniformity=sum(nh.^2);
    texture.entropy=sum(nh.*log10(nh));    
    texture.skewness=skewness(roih);
    texture.correlation=sum(nh'*roih)-texture.mean/texture.std;

    yy=[texture.mean,texture.globalmean,texture.std,texture.smoothness,texture.uniformity,...
    texture.entropy,texture.skewness,texture.correlation];
    y1=[y1;yy];

    xy=[geometric.area,geometric.peri,geometric.compact];
    x1=[x1;xy];

    f2=[x1 y1];
    Totalfeaures=[f1,f1];

[Walter Roberson]

This should have been posted to https://www.mathworks.com/matlabcentral/answers/350414-mammography-image-feature-extraction

[Ankur Pruthi]

Hi mohamed ahmed. Can u please share the code for pectoral muscle removal from the image. It would be of great help to me. Mail id- [e-mail]

Answer 1

[fatin rasyidah]

hi Mohamed Ahmed, may I know what method that u use to extract ROI and segment the ROI. I hope u can share it, because i am currently doing the same thing like u have did. thanks in advance

[Amit manjhi]

Hi Mohamed Ahmed can u share the MATLAB code for Pectoral muscle remove, ROI extraction and Segment the ROI.I am doing the same project for my final year project so can u send plz.It is very urgent. my email [e-mail]

[Samar Qassir]

Hi Mohamed Ahmed can u send me the MATLAB code for ROI extraction and Segment the ROI with thanks my email [e-mail]

[aiesha]

hi to all, i am undergraduate student , and doing work on detecting breast cancer from mammogram images as my project. can anyone please share the MATLAB code for Pectoral muscle remove, ROI extraction and Segment the ROI or any other details or code material related to my project. kindly reply me ASAP . This help will be very meaningful to me.

[Walter Roberson]

http://www.mathworks.com/matlabcentral/answers/29922-why-your-question-is-not-urgent-or-an-emergency

[Hoang Duy]

Hi Mohamed Ahmed can u send me the MATLAB code for ROI extraction and Segment the ROI with thanks my email [e-mail]

Answer 2

[sathya lakshmanan]

Hi Mohamed Ahmed can u send me the MATLAB code for ROI extraction and Segment the ROI with thanks my email [e-mail]

Answer 3

[hamid hosseinabadi]

Hi Mohamed Ahmed plz send to me your MATLAB code [e-mail]

Answer 4

[Loai Kinani]

Hi Mohammad,

can you share the the matlab code of pectoral muscle removal.

Thank you

Answer 5

[ahmed nageeb]

hello every one

when use SegI , give me error

Undefined function or variable 'SegI'.

can help me

[Walter Roberson]

You are correct, the code that was posted does not define several of the variables.

SegI is not a function: it is a variable the user defined but did not show us the construction of.

Answer 6

[mary mwadulo]

hi Mohamed Ahmed,can you kindly share with me the MATLAB code for pectoral muscle removal, extracted ROI and Segmented ROI.Am currently working on the same. my email is [e-mail] thank you.

[manish kumari kumari]

hi Mohamed Ahmed,can you kindly share with me the MATLAB code for pectoral muscle removal, extracted ROI and Segmented ROI.Am currently working on the same. my email is [e-mail]

Answer 7

[Elahe Mgh]

Hi Can you send me the matlab code for extracted ROI and segmented? [e-mail]

Answer 8

[Amelougou Abderrahman]

hello everyone please I need your help urgently...I worked at first with the Matlab FirstCode code which allows me to delete the pectoral muscle after the background deletion but I would like to see my final image of the same original image size that's why I tried to work on the same code with modifications: 1. orientation of the images right to left and 2. do not remove the background 3 .determine the first and the last non-black pixel in the image i.e. I determine my region of interest ROI 4.application of pectoral muscle removal algorithm .....but unfortunately does not give results requested (muscle removal).....please help