3D volume formation & display using 2D image stack

I have 931x1060x123 image stack (double). This comes from a .lif file and constitutes 3D structures. The stack width, height and depth are 3221.0966 (931p), 3667.4140 (1060) and 615.0984 (123) microns. The voxel size is 3459.8x3459.8x5000.8 and the resolution is 0.2890 pixels per micron. I want to display it as a 3D volume. So far I have done this:

clearvars
close all
clc

addpath        ('C:\zmf\data')
addpath        ('C:\zmf\code')
rootdir      = 'C:\zmf';
cd             ([rootdir '\data'])
data         = dir('*Kidney83.tif');
fname        = data.name;
info         = imfinfo(fname);
load           microscopeSpecs.mat
metadata     = microscopeSpecs;  
metadataKeys = metadata.keySet().iterator();

for i=1:metadata.size()
    key   = metadataKeys.nextElement();
    value = metadata.get(key);
    str   = strfind(key, '|');
    if strcmp(key(str(end)+1:end), 'Begin') == 1
        zmax = str2double(value);
    end
    if strcmp(key(str(end)+1:end), 'End') == 1
        zmin = str2double(value);
    end

      fprintf('%s = %s\n', key, value)
  end

ImageSizeReductionFactor = 1;

I = zeros(ceil(info(1).Width*ImageSizeReductionFactor), ...
    ceil(info(1).Height*ImageSizeReductionFactor), length(info));

for i = 1 : length(info)
    im       = double(imread(fname, i));
    im       = imrotate(im, -90);
    im       = imresize(im, ImageSizeReductionFactor);
    I(:,:,i) = double(im);
    names    = strrep(fname, '_', '\_');

      % Display results
      imagesc(I(:,:,i))
      colormap(jet)
      title([names ', Slice : = '  num2str(i)])
      pause(0.001)
  end

%%Display 3D volume
xlength = info(1).XResolution*info(1).Width;
ylength = info(1).YResolution*info(1).Height;
zlength = abs(zmax-zmin)*10e3;
xvec    = repmat(0:info(1).XResolution:xlength, info(1).Height, 1);
yvec    = repmat(0:info(1).YResolution:ylength, info(1).Width, 1);

[faces,vertices]= isosurface(I);

figure;
patch('Faces',faces,'Vertices',vertices)
grid on
colormap copper

but it does not render a 3D volume correctly using isosurface() and I cannot assign a face color myself. It appears all black plus some interpolation stuff removes some data that I wan to visualize.

However, if I use ImageJ plugin (3D Viewer), I don't have to do anything and it displays the volume correctly:

So I want to display something like this. How can I correlate my pixel values and voxel size information to create voxels such that I get a 3D volume? Or is there a possibility to display the stack correctly like ImageJ plugin did without voxel generation for the whole volume?

Best regards,

Answer 1

[Isaac]

Hey, I'm running into the same problem right now. Were you able to solve this?