How to reshape a single column matrix in the following example?

I have a CT dicom files (512*512*263), I wanted to play with the CT numbers (HU) in the following way: (a) multiply the voxels that have CT number less than 20 by 2 (say, for fun, I want to understand how this works) and (b) multiply the voxels that have CT numbers greater than 20 by 10. I wrote the following code, find below.

In the code I splitted the CT_number into two parts and worked on it, now I want to combine the results such that the new column vector 1D will have the updated values. How can I do that? Shouldn't be hard I guess, but I lost. ANy help would be helpful.

CT_matrix = zeros(512, 512, 263);  % preallocate the image array
info_ct = dicominfo('...01.IMA');
 for p = 1:263
   CTfilename = sprintf('..-%03d.IMA', p);
   CT_matrix(:,:,p) = dicomread(CTfilename); 
 end
 % Convert CT matrix into a column vector to assess the voxel values
 idx = CT_matrix(:);   % idx is a column vector
 CT_number = idx ;

 % voxels that have CT numbers less than 20 (say)
 voxels_have_CT_number_less_than_20 = CT_number(CT_number < 20);
 % Multiply the voxels_have_CT_number_less_than_20 by 2
 new_voxels_have_CT_number_less_than_20 = voxels_have_CT_number_less_than_20 * 2;
 
 
 % voxels that have CT numbers greater than 20 (say)
 voxels_have_CT_number_greater_than_20 = CT_number(CT_number > 20);
 % Multiply the voxels_have_CT_number_greater_than_20 by 10
 new_voxels_have_CT_number_greater_than_20 = voxels_have_CT_number_greater_than_20 * 10;
 
 % Now, I want to create a 1D matrix (by combining new_voxels_have_CT_number_less_than_20 and new_voxels_have_CT_number_greater_than_20)  
 % such that I can see the matrix oprtaion in a single column.
 % Then, I can reshape this matrix to the original CT dimension
 new_CT_map = reshape(.., size(CT_matrix));  % not sure how to complete this!

[blues]

After matrix manipulation, I had two vectors in above example. One is new_voxels_have_CT_number_less_than_20 and other is new_voxels_have_CT_number_greater_than_20. Both of them actually belongs to CT_number (1D column) vector.

My question is how can I combine the results of these new_voxels_have_CT_number_less_than_20 and new_voxels_have_CT_number_greater_than_20, so that I can make a new colmun vector (1D), then finally I can use reshape function to make a original size CT matrix?

[blues]

Yes, how can I convert the updated column vectors back to the same shape as in original one?

Answer 1

[Turlough Hughes]

If you write the following you can directly modify the CT_matrix, no need to separate it out into a column vector:

ind_lte_20 = CT_matrix<=20; % index for values less than or equal to 20.
ind_gt_20 = CT_matrix>20;   % index for values greater than 20.
CT_matrix(ind_lte_20) = CT_matrix(ind_lte_20)*2;
CT_matrix(ind_gt_20) = CT_matrix(ind_gt_20)*10;

[blues]

Thank you