Sunday, August 24, 2014

Research I missed: Radiation doses of the old naked scanners

I'm currently going through the recent paper by Mowery, et al on the effectiveness and security of the "old" (x-ray backscatter) naked scanners. Along the way, I came across this paper by the American Association of Physicists in Medicine (AAPM) [pdf], which was given access to the same machines by TSA and Rapiscan in order to assess radiation doses.

Their findings (in brief):
ANSI and the HPS have issued a standard (ANSI/HPS N43.17 2009) that “applies to the manufacture and operation of security screening systems that are intended to expose humans to primary beam x-rays, gamma radiation, or both.” Our measurements indicate that the effective dose from a single screening exam is well below the screening limit of 0.25 μSv per screening for a general use, full-body scanner.5 The standard also states that the effective dose (computational adult model) shall not exceed 250 μSv over a 12-month period. For our stimated effective dose of 11.1 nSv to a standard man from a single screening, an individual would need to go through more than 22,500 screenings in a year to reach this limit.
Which is to say, the effective dose (akin to an overall radiation dose) is quite small and is, indeed (as TSA and Rapiscan said), well below the safety standards that were already in place for medical radiation exposure. It is also well below the threshold of what has been advanced for acceptable levels of non-medical radiation exposure - specifically for the purpose of security screening, commissioned by the FDA post-9/11 - which largely draws on the medical standards. However, it is not as if there is a quantifiable individual health costs vs health benefits that can be made for mandatory screening as a condition of air travel. Of course, such a calculation can statistically be made in medicine, the cost/benefit analysis can be presented to the patient by a doctor, and medical scans are voluntary. I'm not saying it's unsafe, rather that it ought to be up to the individual to weigh the risks with benefits without having to forfeit such fundamental rights such as freedom of movement and freedom of contract.



Nonetheless, the radiation exposure is very low and no passenger - even a frequent flier - is likely to ever reach the threshold of the existing medical scan standards. Which begs the question - why didn't the TSA or Rapiscan allow independent testing prior to 2013? To throw the worn-out adage back at them: If they didn't have anything to hide, they shouldn't mind the extra scrutiny!

So, how does this relate to the earlier simulated study published by Hoppe and Schmidt? If you recall, Hoppe and Schmidt used the few public backscatter images that were available to predict what the radiation dose must be in order to gain the resolution shown in the images. [UPDATE: I did not recall, apparently. The research I described was that of Kaufman and Carlson. Hoppe and Schmidt, rather, ran simulations to determine organ exposure during x-ray backscatter scanning.] Their results were roughly 3-6 times higher than the measurements made by the AAPM group. Here's some reasons why this may be:
The relatively large difference between the total effective dose estimates for Hoppe-Schmidt and the other studies are presumably due to assumptions related to the Monte Carlo models used (PCXMC vs. GEANT4), how the x-ray source itself was modeled, and how the passenger’s body was modeled (standard MIRD computational phantoms vs. voxelized patient models).
As with the Hoppe and Schmidt study, the AAPM group also estimates doses to various organs (in particular, skin, breast, and thyroid). Of course the skin dose is higher than the effective dose, but they also concur that the radiation penetrates below the skin into other organs. This is in contrast to what the TSA has said in their publicity (if not per se in the research they commissioned or cite). And the safety standards for organ doses are on much shakier ground than effective dose.
As stated in the European Commission’s Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR): Health Effects of Security Scanners for Passenger Screening (based on x-ray technology) in 2012: short-term deterministic effects of tissue damage cannot occur at doses delivered by security scanners. Long-term stochastic effects such as cancer risk are assumed to be directly proportional to received dose with no safe threshold.
But, AAPM goes on to say quote that SCENIHR report as saying the cancer risk is negligible for an individual, but can not be truly ignored at the population level. AAPM then concludes with some common sense:
For perspective, we think it important that this potential increase in risk to the population be considered in light of the presumed increase in risk originating from the much greater radiation exposure from the flight itself.
Finally, what about the Mehta and Smith-Bindman article from 2011? Their assumptions did in fact use higher radiation doses than AAPM measured in this study.

I can't disagree with the logic or the statistics of these concluding paragraphs. I just think each individual should be able to make this assessment on their own, not put through mandatory screening. And that is why I was so hard on Mehta and Smith-Bindman - they were looking at the masses and making a recommendation instead of encouraging individuals to make educated decisions.

And, now we are full circle. The reason the TSA and Rapiscan did not allow independent testing sooner is because they are incapable of thinking of the public as thinking individuals with express civil rights. We are a conglomerate who can not be trusted with facts, but must be shaped with propaganda.