Don't know how it's implemented in reality, but I'm guessing it records the diffraction pattern from the blood cells (the hologram) and then guesses the image on the slide by using something similar to a phase retrieval algorithm (http://en.wikipedia.org/wiki/Phase_retrieval). It's similar to calculating the widths of the double slit experiment knowing only the diffraction pattern.Not hard but takes a lot of computation power. We use it all the time in our adaptive optics setup.It's cleaver because he uses cell phone imager that is not really suited for something like this.