Hacker News new | comments | ask | show | jobs | submit login

More tellingly, on an average two-hour flight, you'd receive about 100 times the radiation from a single scan.

But, by using the transitive property, you're more likely to die from radiation at flying altitudes than from a terrorist attack on said airplane.

EDIT: Sorry, I forgot to include the link:

http://en.wikipedia.org/wiki/Backscatter_X-ray#Health_effect...

Which comes from:

http://hyperphysics.phy-astr.gsu.edu/hbase/nucene/radexp.htm...

Which explicitly states its references.




This comparison is meaningless.

The risk of cancer is caused by DNA errors. Your cells are constantly checking and repairing your DNA for mutations and damage. They can usually repair or detect a low error rate, like the kind you get from living at high altitude, in a house with a minor radon source, or flying on a plane. What they can't deal with as well is a sudden burst of many errors. Backscatter X-ray scanners deliver their entire dose to a tissue region in a few microseconds. That's the kind of exposure we're talking about, and it has an entirely different biological impact.


>What they can't deal with as well is a sudden burst of many errors.

Actually that is false, as far as low doses:

>When you are asked whether there is a critical time period over which 1 rem of dose may have a greater biological impact than it might otherwise have, the answer is "No." One rem of dose is sufficiently low that whether it was delivered within one second or spread over a year or more, we would not expect any difference in biological effects.

https://hps.org/publicinformation/ate/q8325.html

(For comparison, 1 uSv = 0.1 mrem).

>The risk of cancer is caused by DNA errors. Your cells are constantly checking and repairing your DNA for mutations and damage. They can usually repair or detect a low error rate

But in this dose regime, ionization events are so rare, that they are so far apart that no cell experiences more than one. There is no variation in error rate.


Good point uvdiv: if the ionization events are sufficiently rare in both cases we ought to be able to treat them as cumulatively equivalent. I guess it would depend on the error correction time and the size of the affected nucleotide sequence, especially in a double-break error. The dose delivered by these scanners is probably well-below that threshold.

My remaining concern would be that the biological effects of this particular energy regime are as-of-yet uncharacterized.


Let me quote from your own source:

"There have been numerous biological experiments conducted, with nonhuman organisms, that demonstrate that the rate at which radiation dose is delivered can affect the extent of biological response. Thus a sufficiently high dose delivered over a period of a few minutes may be expected to have a greater biological impact than the same dose spread over a year."

The only question is what constitutes a "sufficiently high" dose, due to extrapolation. See 5).

THEIR OWN POSITION PAPER on backscatter scanners recommends per-screening and per-year (time limit!) maximums FAR below the threshold that they claimed wasn't a problem in the first place -- see 8).

4) Further, the terse "no" answer looked disingenuous. The cited position paper link was broken but I tracked it down.

It appears to be "Radiation Risk in Perspective", PS010-2 (which I assume is the updated version of PS010-1).

It has nothing to say about doses over short time periods.

http://www.hps.org/hpspublications/positionstatements.html

5) The issue seems to be that there is very little data about low-dose exposures, so regulations are based on extrapolation from high-dose exposures.

See http://www.ncrponline.org/PDFs/TST_NRC%20_04-08-08.pdf (third slide)

6) What it does say is this:

"the Health Physics Society recommends against quantitative estimation of health risks below an individual dose of 5 rem in one year or a lifetime dose of 10 rem above that received from natural sources."

7) The whole point of their objection was that we lack the empirical evidence to extrapolate from high doses to low doses. Then the article you cite does exactly that. They go from saying, "we lack proof that the answer is X" to saying, "the answer is ~X".

8) The recommended standard "limits the reference effective dose delivered to the subject to 0.25 microsieverts (25 microrem) per screening. Additionally, a screening facility should not expose any individual to more than 250 microsieverts (25 millirem) reference effective dose in a year."

"Use of Ionizing Radiation for Security Screening Individuals" PS017-1

http://www.hps.org/hpspublications/positionstatements.html

How they can go from saying in the first paper, "However, below 5–10 rem (which includes occupational and environmental exposures), risks of health effects are either too small to be observed or are nonexistent" to supporting a maximum of 0.025 rem per year is entirely unclear.

Except that the Health Physics Society seems to have turned into a pro-industry spokesgroup.

The current president works for Dade Moeller, who are embroiled in a number of controversies, to put it lightly.

A lot of extremely damaging information is cited in the original link below.

""Hirsch then zeroed in on Dade Moeller’s radiation plan and on the controversial company itself, whose namesake testified back in the 1990s that money spent on cleaning up Cold War-era nuclear facilities was being wasted since there would be a cure for cancer."

Original comment:

http://news.ycombinator.com/item?id=1889673


>THEIR OWN POSITION PAPER on backscatter scanners recommends per-screening and per-year (time limit!) maximums FAR below the threshold that they claimed wasn't a problem in the first place -- see 8).

>...How they can go from saying in the first paper, "However, below 5–10 rem (which includes occupational and environmental exposures), risks of health effects are either too small to be observed or are nonexistent" to supporting a maximum of 0.025 rem per year is entirely unclear.

Where they support that 0.025 rem/year recommendation, they explain that the point is to keep "individual doses as low as reasonably achievable (ALARA) while achieving the desired objective." (Same source, PS017-1).

And you mischarcterize their claims. HPS does not claim <5 rem doses are "not a problem" -- they said the effects are such are "either too small to be observed or nonexistent". (It does not follow that a health effect too small to be epidemologically noticable is "not a problem"). "Too small to be observed" is NOT a quantifier of significance; it is an explanation for the lack of empirical observations. That position statement (PS010-2) is a recommendation against quantiative estimation of the effects of <5 rem radiation doses, not an assertion of their absolute safety.

>It appears to be "Radiation Risk in Perspective", PS010-2 (which I assume is the updated version of PS010-1). It has nothing to say about doses over short time periods.

It does not appear to have been referenced for that purpose.


> It does not appear to have been referenced for that purpose.

Your comment referenced it for that purpose.

Your comment about dose per unit time not mattering referenced an article which referenced their position paper which didn't support what the article said. Hence the article was simply the opinion of the author.

> (And you mischarcterize their claims. HPS does not claim <5 rem doses are "not a problem" -- they said the effects are such are "either too small to be observed or nonexistent". (It does not follow that a health effect too small to be epidemologically noticable is "not a problem").

"Nonexistent" sounds like "not a problem" to me.

"Too small to be observed" is the same. How is a problem that causes no observable effects a problem? That's the practical definition of "not a problem".

And if dose per unit time doesn't matter, then their per-screening and annual limits for these small doses wouldn't matter. They are contradicting themselves.

Further, as mentioned, they are not an independent safety group. They are an industry group for the pro-nuclear lobby. This is like relying on the Business Software Alliance (BSA, a front group for Microsoft and other large software corporations) for opinions on the harmful nature of DRM, copyright laws, etc.


Where 'an entirely different biological impact' may just as well mean 'no impact at all': it may even turn out to prevent melanomas.

While everyone is busy losing themselves in radiation hysteria, where facts are only guessed at, the government is laughing their ass off, because everyone lost sight of the actual problem: that these machines are assumed to improve safety in the first place. The radiation scare is an enormous red herring that mainly harms the public perception of 'radiation', that wasn't too good to begin with, for all the wrong reasons.


We know that short-term small doses of X rays results in increased probability of developing skin cancer. We know that prolonged exposure to low-level radiation sources increases your probability of developing skin cancer. We know that acute high doses of ionizing radiation leads to your skin falling off. Skin cancers are abnormally elevated in radiologists, survivors of Hiroshima who experienced even relatively low doses, radioisotope miners, and people who are treated with X rays for skin conditions.

I'll hazard a guess here and say there's close to zero probability that backscatter machines are lowering your risk of cancer.

Don't get me wrong: I love radiation. I think Cherenkov glow is the coolest thing about working in a reactor. I've handled more sources than I probably should have by this point in my life. At the same time, we should be honest that there is probably a small risk, and it is probably higher than currently stated.


  We know that short-term small doses of X rays results in
  increased probability of developing skin cancer.
An increase that is so minute that we don't mind the dentist or doctor X-raying us. An increase that is so minute that we would accept it, if it actually reduced our chances of dying due to a terrorist attack. The radiation is not the problem: it is the intended goal that we have no reason to believe in.

  We know that prolonged exposure to low-level radiation
  sources increases your probability of developing skin
  cancer
Do we? Not WiFi or the low-energy spectrum of the sun. What energy is 'low' in this respect?. And how long is 'prolonged exposure'? These claims are undoubtedly true for some values of the aforementioned quoted terms, but are they relevant here? Do most of the people repeating this stuff even wonder about that? Again, the problem is a slightly increased risk. But why is that increased risk a problem in the first place?

  I'll hazard a guess here and say there's close to zero
  probability that backscatter machines are lowering your
  risk of cancer.
And I agree. But the "we don't know for sure" argument is pretty shallow, when we accept certain risks on a daily basis and when we all know certainty doesn't exist and we all take aspirin to prevent a second heart-attack, even though a thousand doctors will argue with another thousand doctors about whether that actually works, all citing studies supporting their side, and the net effect is quite small even according to the side affirming it works.


Preface: I totally agree with your argument that risks must be weighed against outcomes. I also believe that even if the scanning technology were totally safe, it would be objectionable on privacy grounds. I think we're going about the whole problem of security backwards. :)

An increase that is so minute that we don't mind the dentist or doctor X-raying us.

Actually, we do. There are good reasons we impose strict limits on your medical X-ray exposure annually, and why you wear a lead vest or use collars to protect other tissue. There's also a good reason the radiologist stands outside the other room to press the button. Two x-rays a year is not significantly mutagenic. Fifty? Then you start getting nervous.

Not WiFi or the low-energy spectrum of the sun. What energy is 'low' in this respect?. And how long is 'prolonged exposure'

UV exposure from solar radiation is one of the biggest melanoma risks. A few orders of magnitude up the frequency scale, X rays for far shorter durations also have mutagenic effects on skin tissue. It's possible that low-energy X rays, falling in the middle of that energy range, also cause cancer. We don't know for sure yet because we don't regularly expose people to them.

(I should mention that just because the X rays are low energy does not mean their effects scale linearly. Medical X rays are well above the Compton shoulder. IIRC (not enough time to check now) they're thermalizing slowly in heavy tissues as opposed to bouncing inelastically off the surface layers. Depending on the scattering cross-section for those energies, the backscattered X rays could be more ionizing. Total conjecture... I haven't done this for a while.)

Again, the problem is a slightly increased risk. But why is that increased risk a problem in the first place?

You're absolutely right. We have no good characterization of the risk. Radiation biometrics is fucking hard. I only want to argue that our current risk statement is inaccurate and poorly understood (for instance, we don't even know the angular flux of these machines), and offer some potential reasons why the risk might be higher than currently stated.


You are understating the risk from backscatter radiation. Because the radiation is sent in a burst, and targets skin tissue, the DNA damage may be far greater than a similar dose of cosmic radiation.

You are also assuming that the backscatter machines are functioning properly and aren't overdosing people.


  Because the radiation is sent in a burst, [..] the DNA
  damage may be far greater than a similar dose of cosmic
  radiation.
People keep repeating that, but it is not simply true. DNA can get damaged when a single photon excites an electron and consequently influences the molecular structure. The chances of multiple photons operating on the same electron at the same time, even in cases of a radiation burst, are negligible. The person is affected by the same number of photons, which results in the same number of chances of electrons being excited. The number of excited electrons will be equal and the resulting damage will be equal.

Now it may matter that the damage is more localized in both space and time, but no one has shown this. Firstly space: spreading the radiation over only the surface, for instance 1/10th of the body (remember that surface times a small depth is always a relatively large part of a volume), means 10 times as much damage. But if the average amount of damage in the skin was one broken bond per 1000 cells, it will now be one broken bond per 100 cells. As the repair takes place per cell, that doesn't change anything for the total chances to get cancer. Only if the amount of damage in a single cell, and close enough to each other to be relevant, changes, then this type of operation may pose a problem.

Secondly for time: everything depends on the amount of time it takes a cell to repair DNA damage. If the average time of repair takes 3 seconds, then during a flight, some damage will be repaired before other damage takes place. The backscatter machine delivers its dose faster and damage may accumulate. However, if repair takes 30 minutes on average, the difference is again neglible.

The point is not that I know the numbers: the point is that people making claims don't know them. They are just claiming things that seem to make sense to them, when they draw an analogy with a car crash or something. They don't understand what is actually involved.


A single backscatter X-ray scan gives a radiation dose of 0.009 mrems, about the same as eating a banana.

http://en.wikipedia.org/wiki/Banana_equivalent_dose

Don't get me wrong, I'm opposed to the scanners as much as anyone for privacy reasons and general pointlessness. But I'm seeing a lot of bad science thrown around here.


You're not comparing bananas to bananas.

As I understand, the limited information that has been publicly released about the backscatter machines is a useless measurement when you take into account how they actually work. You can't apply a standard measurement from a medical x-ray machine that targets a small part of your body using one technique to a different machine that targets your entire body using a different technique.

So the two problems are:

1. You don't have good numbers to even evaluate how dangerous they are, many evaluations, including possibly the article above, are useless because they don't have the real data to compare to.

2. The radiation exposure from a banana also effects your body differently than a medical x-ray machine which is in turn different from a backscatter machine targeting your entire body.

I'm not saying people aren't trying to evaluate the actual risk in good faith. I'm saying they're skipping the step where we get useful numbers from the manufacturer of the machine to even apply good science to.

Edit: by "whole body" I mean "whole surface of your body" since it's concentrated on your skin


While I might agree that the worried about an X-ray scan are overblown (I assume that is your position), that number is looking at a whole body number. Part of the argument is that the amount may be significantly higher because the dose is concentrated near the surface of the skin, rather than across the whole body like cosmic ray and chest X-ray exposure.


So instead of being equivalent to eating a banana, it's probably equivalent to mashing one up and using it as skin cream.

In other news, millions of small children who enjoy playing with food exposed to RADIATION! Are bananas safe for your children?




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: