3d image of a computed tomography

I have many shots from a computed tomography. I have used the matlab patch function to plot a 3d image. Here is the code:

      data = smooth3(data);
      p = patch(isosurface(data),'FaceColor','yellow', 'EdgeColor', 'none');
      isonormals(data,p)
      view(3); axis vis3d tight
      camlight; lighting phong

There is a spacing of 2mm between each plane of the tomography. The spacing of x and y axis is 1mm. I think the voxel must be rectangular. How can i do it?

Answer 1

[Sven]

The isosurface function can take in an x, y and z vector specifying the location of each voxel. So try this:

xVec = 1:size(data,2);  % "X" is the 2nd dimension of the matrix
yVec = 1:size(data,1);  % "Y" is the 1st dimension
zVec = (1:size(data,3))*2;  % Multiply by 2 for 2mm spacing

p = patch(isosurface(xVec,yVec,zVec,data),'FaceColor','yellow', 'EdgeColor', 'none');

Does this work for you?

You may also have your exact X, Y, and Z spacing defined in mm, embedded in the .PixelSpacing (for x,y) and .ImagePositionPatient (for each slice of z) fields of the original DICOM files.

Answer 2

[SoyLeyenda]

Thank you for your answer Sven. The image is longer now but it seems to loose precision.

[Sven]

As long as your actual pixel (well, "voxel", because it's 3d) dimensions are true to the original voxel space, then I'm afraid that you didn't _lose_ precision at all... it's just that your Z precision was *always* lower than your X-Y-precision. Is the question answered or are there still any issues you're having?