Calculating percentage of brown pixels over other colors in tumor

Hi everyone,

I want to calculate percentage of white signals over black ones (but just in the tumor roi). I mean the code that I am using now calculating all black pixels in picture so gives me small percantage but I need calculate just in the tumor area.

Can you help me to remove other black parts from calculations?

THANKS

Answer 1

[Simon Chan]

Define the tumor roi and then calculate the white pixels inside the roi:

clear;clc;
rawdata=imread('image.png');
BW = rgb2gray(rawdata);                                             
J = imclearborder(BW);                                              % Clear pixels in the boundary
black = sum(BW==0,'all');
white = sum(BW~=0,'all');
tiledlayout(1,2,'TileSpacing','none','Padding','none')
nexttile
imshow(rawdata,[]);
title(sprintf('Original Image, Black Pixel: %d, White Pixel: %d',black,white));
BW2=bwconvhull(J);                                                  % Finding the mask
BW3 = bwperim(BW2);                                                 % Finding the perimeter, only for figure display
[r,c]=find(BW3);                                            
for k = 1:length(r)
    rawdata(r(k),c(k),1)=255;                                       % Set the pixels in the perimeter to red
end
black = sum((BW==0).*BW2,'all');
white = sum((BW~=0).*BW2,'all');
nexttile
imshow(rawdata);
title(sprintf('Mask Region, Black Pixel: %d, White Pixel: %d',black,white));

[Ugur Akca]

Thank you so much for your help, it is working great and right now,my results is more reasonable.