I am not sure I understand what you would like to achieve exactly, but here are a few suggestions:
If you have p images, each of size n-by-m, why not simply store them in an (n x m x p) matrix?
What do you mean by 4-D array? What does each dimension correspond to? If you want to store, for each pixel, the 3-D position and the intensity at that position, you could have a simple N-by-4 matrix (2-D): the three first columns are for the x, y and z positions and the fourth column is the intensity value.
You won't be able to get the intensity at an xyz position from the 3-D plot directly. To link intensity values with xyz positions, you would have to find the transformation from pixel locations in the image space to the 3-D coordinates, as defined by xPicture, yPicture, and zPicture.
If you have some data we can play with, that would be easier to help you.