In the recent weeks TSA started to aggressively steer people towards the whole-body scanners, which are capable of producing "naked" images of people. This policy was introduced without much public debate, raising numerous concerns about privacy, legality, civil rights, etc. In this article we'll concentrate on safety of these screening devices.
TSA and a number of officials from FDA have issued assurances that these scanners are safe, claiming that a number of experts have reviewed the radiation exposure data and agrees that the doses of radiation travelers get from being scanned are well within exposure limits established as safe. However, the technical specifications, details of operation and construction, and other data necessary for an independent review of safety of these devices are not published.
What is known is that there are two different types of scanners — one uses scattering of "soft" X-rays, and another uses the terahertz (millimeter) microwaves to form an image. We will discuss the X-ray scanners first.
The general public is well aware of danger of exposure to X-rays; this danger is being forcefully underlined by the usual safety procedures of clinical radiologists donning lead vests and retreating before turning the X-ray machines on. Unlike high-intensity radiation, low-intensity X-rays do not kill or burn cells outright, but the energetic photons can and do damage DNA in the cells. Most DNA damage is not critical, and is repaired by the cells; however some damage remains unrepaired — crippling cellular mechanisms, or disabling them completely when enough damage is accumulated. One of these mechanisms, called apoptosis, or programmed cellular death, prevents cells in our bodies from multiplying uncontrollably. Once this mechanism is disabled, the cell becomes cancerous. (Radiation exposure was also recently found to increase mortality from cardiovascular diseases).
It is important to understand that because of this cumulative nature of damage caused by radiation, the effects of being exposed to additional radiation do not usually appear immediately. It often takes 10-15 years from the initial exposure for the irradiated tissue to become cancerous — by which time it would be impossible to say what exactly caused the cancer. The dangers are better understood in terms of number of unnecessary deaths the scanners would likely cause given the millions of travelers and crew members being scanned every year.
The commonly used estimate is that relative risk of death from cancer increases 3% for every 10 mSv or 1000 mrem of additional exposure (J.P. Ashmore et al, First Analysis of Mortality and Occupational Radiation Exposure based on the National Dose Registry of Canada, Am. J. Epidemol. (1998) 148(6): 564–574) — with melanoma (an aggressive and usually fatal form of skin cancer being the biggest contributor). For comparison, a single cross-country flight on an airplane yields about 4 mrem of exposure, and chest X-ray is at about 10 mrem.
TSA claims that single whole-body backscatter scan yields only 0.002 mrem of exposure. For now, we will accept this figure.
What TSA and FDA experts do not say is that X-rays used by the scanners are very different from the X-rays used in the medicine. Medical X-rays are "hard," and use higher frequency photons with energies from 50KeV to 150KeV. These energetic photons are penetrating (like background radiation from the Sun, radioactive elements in soil, etc), and for them biological tissues are semi-transparent. The "soft" X-rays used in backscatter machines (20KeV or less) are mostly absorbed in the skin — making them useless for medical imaging. All medical X-ray emitters are equipped with special filters to suppress soft X-rays — both to increase clarity of images and to reduce exposure.
The health effects of hard X-rays are well studied due to their wide-spread usage. Information on health effects of soft X-rays is scarce.
Because of the penetrating nature of hard X-rays, the exposure is spread all over the body (and exposure limits are calculated per kilogram of body mass). With soft X-rays the exposure is concentrated in the skin, which makes the effective exposure per unit of mass of tissue 100 or more times bigger than for hard X-rays.
Unfortunately, skin is exactly the wrong tissue to gather the additional radiation exposure. Unlike muscles, fat, bones, or most internal organs skin contains very large numbers of quickly dividing cells. We are constantly growing the new skin and hairs from inside, shedding the old dead cells on outside. The abundance of dividing cells means that DNA damage is quickly multiplied by getting into daughter cells, and that some mechanisms which suppress cell growth are less active in skin cells. These are the reasons why skin cancers are the most prevalent kind of cancer (though the mortality is usually low except for melanomas).
With penetrating X-rays, most absorbed X-ray photons are absorbed in dense tissue and bones, which are much less pre-disposed to cellular damage turning into cancer. This may multiply the biological effects of soft X-ray irradiation by as much as an order of magnitude.
Another difference between soft and hard X-rays is their absorption by the tissue. Absorption depends critically on the probability of interaction between particles (such as photons) and atoms, a property called "cross-section" by the physicists. You can understand cross-section by imagining atoms to be targets and photons to be bullets with cross-section being the size of the targets. Perhaps counter-intuitively, cross-section is often bigger when particles have smaller energies (this is the reason nuclear reactors employ moderators to slow down neutrons so the cross-section of their interaction with uranium nuclei will increase). Similarly, soft X-rays are more likely to be absorbed by the tissue (and thus cause damage) than the hard X-rays. The mass energy absorption coefficient in soft tissues for 10KeV X-rays is 4.6 cm2/g versus 2.5 cm2/g for 100KeV (NISTIR 5632, Tables of X-ray Mass Attenuation Coefficients and Mass Energy-Absorption Coefficients…, J.H.Hubbell and S.M.Seltzer, 1996) — meaning that soft X-ray photons are absorbed nearly 2 times as often as hard X-ray photons.
Taken together, these equivalent-dose multipliers can amount to a factor of about 2000.
One more issue to consider is intensity of X-rays. Medical imaging uses essentially point-like sources, with rays spreading out from a bright dot on a tungsten electrode, resulting in a uniform low intensity. Although the details of construction of the backscatter whole-body scanners are not published, it is possible, by simply looking at the geometry of these devices, to conclude that they use a moving spot of concentrated X-rays scanning over the body. While the overall dose could be small, the intensity (brightness) in the spot can be high.
The probability of DNA damage depends significantly on the X-ray intensity. DNA is two-stranded, and damage to one strand is relatively easily and reliably repaired by the cells because the other (complementary) strand serves as a template for repair. Damage to both strands in the same or nearby locations is much harder to repair — there's no longer a reliable second copy. The higher intensity of X-rays, the more probability is that two photons will strike DNA molecule near each other.
The simplified rules radiologists use to calculate exposure limits do not take into account this intensity dependency because most medical X-ray sources use similar intensities, just sufficient to provide sufficient exposure to the films or sensors. Applying these rules to estimate safety of exposure to higher intensity X-rays is simply not valid.
All these effects can make biological effects of radiation from the whole-body scanners to be up to 10000 times stronger than effects of the same dose of X-rays from a medical X-ray source — i.e. the equivalent effective dose of radiation would be more like 10–20 mrem per scan, using the published scanner radiation exposure figure.
However, this starting figure, 0.002 mrem, is highly suspicious. To create an image, a sensor has to capture sufficient energy (i.e. to have sufficient exposure) to "heat" pixels above thermal and other noise in the sensor (i.e. it has to have a good signal/noise ratio). It does not matter if radiation passed through or was scattered. Are we being led to believe that manufacturers of the backscatter X-ray scanners have invented some revolutionary new X-ray detection technology which is 5000-10000 times more sensitive than the sensors used in mere medical devices? Why these fantastic new sensors are not used in medicine then, to reduce doses to the patients and doctors? Would they care to explain how did they manage this minor miracle?
Assuming that there was no such break-through in sensor technology, the 0.002 mrem figure is simply a lie. (Alternatively, this could be simple incompetence — such as using X-ray radiometer designed for calibration of medical X-ray sources to measure exposure in a full-body scanner. However, the first thing a medical radiometer does is filtering out soft X-rays and other electromagnetic radiation with irrelevant frequencies, so that using such radiometer in a backscatter scanner will result in readings which are way too low).
Given so little actual information on the construction of these X-ray scanners, it is impossible to estimate the actual level of danger they pose to the public; but until all the concerns above are addressed in a credible way, we should assume that the backscatter X-ray scanners are at least as dangerous as medical X-ray machines — and probably more.
Using the increased risk per dose estimate above and multiplying that by more than 750 mil of passengers per year (assuming that all of them are scanned once, and using the generally accepted linear risk dependency per dose of radiation) we can estimate about 225 thousand of additional cases of cancer per year (3%/1000 mrem * 10 mrem * 750,000,000) in addition to 2 million cases diagnosed in US yearly (Skin Cancer Facts); resulting in approximately 1700 additional deaths. This makes these machines about six times more dangerous than terrorists: the average number of people killed in terrorist attacks is only about 286 per year for the 11 year period spanning 1995 through 2005.
If the manufacturers of these machines are lying about the radiation exposure, the danger to the public can be much higher — up to 500 times more, if we assume that sensitivity of the X-ray sensors used is in line with the common medical sensors. Deployment of these machines today could become public health disaster ten years later.
The terahertz microwave scanners are a bigger unknown, mostly because the ability to produce terahertz radiation with useful intensity is relatively new. The actual numbers on intensity of terahertz microwaves emitted by these scanners is not published. Only little is known about the health effects of terahertz microwave radiation. TSA and FDA officials mislead public by comparing this radiation to more common (and relatively benign) gigahertz microwaves used in cell phones and wireless data networks.
What is different about terahertz radiation is that it was shown that many biological molecules, including DNA, have resonant frequencies in this range. (Resonance is a phenomenon known to anybody who ever pushed a playground swing — when you get the frequency of pushes and pulls just right, it really starts to swing). When photons have just the right frequency for a particular kind of molecules, these molecules start absorbing energy from these photons. Non-resonant microwaves simply heat the tissue; hit the resonant frequency, and they start tearing the specific molecules apart.
How much this is dangerous to the live organisms for the intensities and specific frequencies of terahertz radiation being used is largely unknown. In fact, it won't be possible to learn about these effects until 10–20 years worth of statistics is collected. Essentially, US travelers are being forced to become unwilling and mostly uninformed laboratory rats. Same is true about effects of small doses of soft X-rays.
We should understand that claims of safety made by FDA and TSA are not based on any kind of empirical evidence — we will not have this evidence without clinical trials taking decades; the basis for their assurances is opinions of their "experts." We have shown above that these opinions are not grounded in science, and are merely the result of mechanical and scientifically invalid application of safety data from other frequency ranges — despite the obvious differences in biological effects.
We should demand removal of these machines before the actual trials establishing their safety empirically are done, and before the details of construction of the specific models are made public and available for independent review. There are sufficient grounds to challenge the validity of the theoretical numbers used as a basis of claims that these machines are safe. The valid trials cannot be completed in less than 10 years — simply because it takes that long for cancers to appear following the exposure.
Meanwhile, travelers would be well advised to stay clear of the whole-body scanners.
Vadim Antonov [send him mail] is an entrepreneur, software and networking engineer, and security expert. He co-founded the first Internet Service Provider in the USSR, and was one of the first to recognize value of the Internet as a political tool, using it to help overthrowing the Soviet communist regime in 1991.