reconstruct 3D image

Hi, i have 20 of 2D images in dicom format and i want to reconstruct the 3D image. How can i do this in MATLAB? here i attached my code and i hope anyone could help me to fix it. Many thanks

a = dicomread ('brain_001.dcm')
mask = zeros(size(a));
mask(100:end-50,100:end-50) = 1;
bw = activecontour(a,mask,1000);
% for p=1:20
%    filename = sprintf('brain_%03d.dcm', p);
%    a(:,:,1,p) =double(dicomread(filename));
% end

 c = im2double(bw);
 shading flat

 d = imgaussfilt3 (c,2);

 for p=1:20
   filename = sprintf('brain_%03d.dcm', p);
   a(:,:,1,p) =double(dicomread(filename));
end

 colormap(copper)
  e = mesh(d.*100)

[Image Analyst]

What do you mean by "reconstruct"? And why are you putting the 2D images into a 4D array? And why are you computing "c" and "d" but not doing anything with them?

[Walter Roberson]

implay looks for M-by-N-by-1-by-K for grayscale.

[Zaitul]

i want to make the 3D image from those 2D images. what should i do for c and d? i'm a bit confused when i'm doing the loop for p. i'm sorry as i'm just realized there's only 20 slices of image actually.

Answer 1

[Ahmet Cecen]

Too vague to answer directly, but here are two helpful tips:

1) Whatever you are doing, in the end you need to stack the images on the 3rd dimension not 4th.

NOT

a(:,:,1,p) =double(dicomread(filename));

BUT

a(:,:,p) =double(dicomread(filename));

2) Do not try to visualize 3D data in MATLAB, unless it is a surface. MATLAB figures (as far as I know) do not support pixel shading, so your figures will look very flat. Instead use Paraview. You can use this function to export your 3D matrix once you have that in MATLAB. Below is an example of a 3D reconstruction I visualized following the same procedure.Notice the depth/shading.