Masking dicom multiple files with same mask

Dear all,

I want to mask dicom files (#slices>100) with same mask and than do the histogram for all slices of the unmasked part and masked separately . For single (1 slice file) it works perfect:

test_image=dicomread(fullfile(pathname, filename) );
imagesc(test_image)
test_if=imfreehand;
notmask=createMask(test_if);
mask=~notmask; //since I want to have everything except mask area
test_image(mask);
test_image(~mask)=NaN;
imagesc(test_image)

Histogram also works:

   hist(double(test_image(:)).

But I can't find the right indexing method for masking all slices. I tried:

for cnt=1:p X(:,:,1,numel(dicomlist))=dicomread(fullfile(pwd, dicomlist(cnt).name)); % Xmask(:,:,1,numel(dicomlist))=double(X(:,:,1,numel(dicomlist))).*mask; % Xmask(find(Xmask==0))=NaN;

end (the upper sometimes work)

or in a different way, similar to single file (below differen tries):

Xmask(mask,1,numel(dicomlist));
Xmask(mask, numel(dicomlist);
Xmask(:,:,1,p)=X(mask);
That all failedm, usually with an error that indexing has to be placed in the end or there is dimension mismatch.

So I am both curious and confused..What is the proper way of doing that? Thank you in advance!

Answer 1

[Henric Rydén]

Hi,

you should load all your dicom files into a 3D matrix. Then, make your mask 3D using repmat. Your code should look something like this: (I haven't tried it so expect errors)

slices = 50; % I don't know how many slices you have, change this
test_image=dicomread(fullfile(pathname, filename));
myAwesomeData = zeros(size(test_image,1),size(test_image,2), slices);
myAwesomeData(:,:,1) = test_image;
test_if=imfreehand;
mask=createMask(test_if);
test_image(mask)=NaN;
figure;
imagesc(test_image)

for idx = 2 : slices
myAwesomeData(:,:,idx) = dicomread(dicomlist(idx).name);
end

maskedData = myAwesomeData(repmat(mask,1,1,slices));

[Aga]

thanks a lot! yes, indeed repmat did the job!

[Stelios Fanourakis]

Excuse me. To this line myAwesomeData(:,:,1) = test_image;

It gives me the error 'Assignment has more non-singleton rhs dimensions than non-singleton subscripts'

what does this mean?