Backscatter hubbub

There’s been a lot of hubbub and commotion in the news recently about the backscatter security scanners (Advanced Imaging Technology) being introduced at airports. You’ll hear all kinds of claims from both sides about them, some partly true, mostly hyperbole and FUD.
One of them involves the radiation dose to the skin. There are claims that the energy is deposited mostly in the dead layer of the skin, or that the radiation dose is very minimal (from the reports I’ve read, it’s on the order of a few μSv). On the other side you have claims that since all of the dose is deposited in the skin, the risk of skin cancer is greatly increased (not entirely accurate and a bit of an exaggeration at the doses involved).
While I was skimming through all of these discussions, the thought occurred to me that I could use MCNP to do a quick “back of the envelope” simulation to see just how the radiation dose might be deposited. Although I don’t know the details of how the components of these scanners are arranged, I think I can make a few reasonable assumptions based on what I’ve read so far. A “passenger” could be modeled as a rectangular slab of water and assuming a scanning beam is used, the source could be modeled as a uniformly emitting plane source. This would let me determine depth dose curves pretty easily.
I’ll have to do some literature searching to see what else I can find out. I’ve heard of papers where people have measured the doses but I think those have been limited to just measuring the radiation output without looking at dose distribution.
This could be an interesting little distraction.
Update: In the TSA Reading Room you can find a moderately redacted analysis document prepared by the Johns Hopkins University Applied Physics Laboratory that describes their measurements of the radiation output from an early version of one model of backscatter scanner.