Slashdot Mirror
Visible Light 'X-Ray' Sees Through Solid Objects
disco_tracy writes "Some day we may not need X-rays to see inside people, thanks to a new way to decipher light that passes through opaque surfaces. Normally visible light becomes too scattered to detect after passing through opaque surfaces. But scientists in France have developed a way to reconstruct images from light passing through such surfaces by deciphering just how the material makes the light scatter. In the short term the research will help improve the strength of telecommunications signals and fiber optics cables, but years from now the technology could supplement or even replace traditional ultrasounds for baby imaging and X-rays for weapons detection at airports."
This is a double edged sword. I see good uses, then I see this used to try to execute search warrants, saying that anything in someone's domicile is in "plain view".
Of course the voyeurs will also love opaque viewing technology too.
This will make the "xray" threads on /b/ waaay more interesting.
There's no -1 for "I don't get it."
This is a pretty cool idea, but it will probably not take the place of x-rays. X-ray is cheap, easy, accurate, and relatively harmless (in small doses).
This sounds expensive, requires a large amount of processing capability, isn't very portable, and relies on light actually passing through the object. For some applications this may be useful, but for the vast majority of imaging tasks that require visualizing the internals of an object, x-rays will be the better solution.
Now, an x-ray scanner that didn't require film plates. That would be good!
How does visible light make its way through an opaque object?
I want to delete my account but Slashdot doesn't allow it.
tin foil underwear.
especially the ones that work in the TSA and homeland security. Yet another way for them to perv out on the job.
How does visible light make its way through an opaque object?
I know you aren't supposed to read TFA, but ""It's like putting a flashlight behind your hand," said Sylvain Gigan... "You cannot see an image, but you can still see a faint glow.""
T'is a little thing called density. And there is too much of it.
Mind you, i'm no expert on the density of flesh between sensor and the babymaker, so it could possibly work.
I do remember i used to shine lasers in to the webby parts of my hand to see it come through really vibrant reds.
But then this just makes me think the opticians are trying to make weak-sighted babies. Conspiracy!
That's exactly what I thought. It's a poor choice of words, in my opinion. Opaque by definition means that it blocks light from passing through it, but I just figured it was some kind of quantum mechanical thing, just like all the other physics I don't understand.
Funny may not give karma, but +5 Informative never made anyone snort coffee out their nose.
No. Reading is useless without understanding. The OP was correct in asking. Your hand is not opaque, it is translucent.
"National Security is the chief cause of national insecurity." - Celine's First Law
X-Ray. See through. Solid. You keep using those words. I do not think they mean what you think they mean.
It's only a model.
I say why stop at x-rays. Gamma rays are too good for them.
From Wikipedia "An opaque substance transmits very little light, and therefore reflects, scatters, or absorbs most of it." Yes, it's making it seem more than it is, but it's still correct (unless wikipedia is wrong, which is impossible).
opaque:
1. Impervious to the rays of light; not transparent; as, an opaque substance.
Translucent:
1. Transmitting light but causing sufficient diffusion to prevent perception of distinct images.
If an object is opaque there is no "light passing through such surfaces" that can be deciphered. It is call opacity but opaque means 0 light pass through.
I know you aren't supposed to read TFA, but "'It's like putting a flashlight behind your hand,' said Sylvain Gigan... 'You cannot see an image, but you can still see a faint glow.'"
I think it would help if TFA included an actual example image, and not just a photo of someone holding their hand up behind a shower screen and a note to the effect that the actual technology might produce images sort of like that one.
"...always new atoms but always doing the same dance, remembering what the dance was yesterday." -Richard Feynman
I guess the first question is whether something can be truly opaque (zero light travels through) or whether all things are translucent if you've a sensitive enough detector.
Assuming that there are genuinely opaque objects, are there enough objects that are translucent (though not to the unaided eye) to make this technique interesting?
My guess is that almost everything will be translucent, though not everything. If the gaps between atoms is on the scale of the wavelength of light, then the atoms will act as a diffraction grating. Given the number of such gratings light has to pass through for any meaningful object, that's going to make a serious mess of the observations.
In order to be truly opaque, two criteria must be met - every photon has to intersect a particle and for every such intersection, the particle has to be able to absorb the photon. Since matter is mostly empty space, you'd need an awful lot of particles to absorb all photons. However, I can see no obvious reason why it would be impossible to have such an arrangement.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
unless wikipedia is wrong, which is impossible
You make me want to go and edit the elephant entry again. ... Or maybe go and randomly edit something I know nothing about basing all my information off of other wikipedia articles, quoting them for authority. I think I might combine something about army ants, satellite antennas, and low-end computer speakers. Or maybe I'll just add the word "not" in front of a significant statement in one of the articles related to a student's upcoming paper to see if they bite the poison apple. Anyone else with me?
Amazing! A friend of mine has done his Ph.D. in exactly this field. He was shining a beam of light right THROUGH an opaque sheet of material (paper, I think) already a few years ago, and published about it in 2008. I think it's pretty much the same idea, from what I understand of it (but keep in mind, I chose the evil path of Business instead of Science, so I have no brain).
Anyway; on his page there's a much better explanation, with cute pictures and all that, of the same idea.
Did it hurt the second time?
Simple, on exit from the opaque object, the light has turned invisible. This new device can see the invisible light.
How about a dictionary, not a wikipedia/"encyclopedia" entry ... "does not transmit light."
I've had this idea for a while now that low-heat, very bright LEDs are available as light sources: 1. take an existing CAT scanner: Xray source, detector, mounting system (with the rotating arm) and image processing software. 2. replace the Xray source with a bank of LEDs 3. replace the Xray detector (a scintillation screen? whatever it is) with a CCD 4. start scanning Obviously there's a whole bunch of experimentation needed to calibrate diffusion due to different types of tissue/bone/marshmallow but the software should be mostly unchanged, the mechanical mounting system would be mostly unchanged, and we'd be replacing a radioactive source with a low-power, low-heat light. Is anybody working on this? I've asked a couple of professor at a biomedical engineering department but much silence ensued. The ability to use off-the-shelf components seems like a big plus to me... There would also be a need to check at what intensity cold light is detrimental to cells (and other small issues like that)
Black holes?
The actual "FA" is here, with images. Gigan, et al. say, "opaque materials."
Credo sim. - I think I am.
Thanks to quantum tunneling nothing is ever completely opaque. A particle's path from A to B doesn't necessarily have to pass through all the points in between. Some tiny fraction of the photons will always act as though the object isn't even there.
"The state is that great fiction by which everyone tries to live at the expense of everyone else." - Bastiat
... see anything more than the silhouette of what is being concealed by the "opaque" surface though? While in many cases, a silhouette could well provide enough for a lot of different purposes, I don't think it's quite what I'd consider really "seeing" something.
File under 'M' for 'Manic ranting'
Take your pick: Opaque, opaque, opaque, either way, his question is valid and the wikipedia entry on the subject matches not only all online and offline dictionary definitions but also textbook ones. And the answer that human hands are translucent is also valid. So give'm a break?
Ever put a powerful flashlight against your fingertips?
... and not just a photo of someone holding their hand up behind a shower screen ...
That picture looks suspiciously like the poster of this movie.
The title says "Visible Light 'X-Ray' Sees Through Solid Objects". So it's a done deal? We have a machine that you stand or sit in front of and we have pictures equivalent in usefulness to a standard x-ray?
(reads summary)
Oh, it's decades away, if at all, for now the application is for one single case where the material is not really "opaque" in the everyday sense...
You know, we complain about shoddy science reporting but aren't you Slashdotters just as bad?
This is almost as bad as a Space Nuttery story: "We can build 5000KM solar wind sails and get 1c /KW/h power!"
Um, no we can't.
2) An opaque object means that light bounces from it, therefore see the opaqur object and we can't see what's in the other side.
3) if the best analogy they could have come up with is a hand with a potent source of light that allows to see the blurry hints of bones and veins of our hand, it is translucency not opacity.
It makes sense to recompose scattered light with algorithms, but it doesn't make sense if they insist calling it "seeing through opacity".
If they claim that they can see through a rock "with visual light" then it is truly something I would like to have next to my IR filters in the beach. (remember the nightshot + Ir filters scandal? Lol)
How does visible light make its way through an opaque object?
Depends the material. Metal foil 0% gets through. Human flesh is mostly water and translucent. Ever hear of a fluoroscope? I question how clear an image will ever be given the different densities of material making up a human body. Say for your head the brain, skull and skin all have different diffusion rates. Even soft tissue will vary. Muscle, skin and fat will have different rates. The process will be more useful in a controlled environment where there is a single material they have to allow for not something as random as the human body.
I hope you are joking around.
Now if someone can just develop a technology to let me throw my voice, my back-of-the-comic-book dreams will be fulfilled.
Do you have a flag?
So in other words, quantum mechanics is opaque to you. Or at best, translucent.
If I have seen further it is by stealing the Intellectual Property of giants.
The idea is that you send light against an opaque medium, the photons getting blocked or scattered is a statistical process. Some of them, simply as a matter of probability, "sneak through" in a straight line.
To get around the low probability, you use a strong light source, modulate it (if you modulate the light, you can pick it out with a tuning circuit, so that you can screen out background light), and then average over a long period of time.
Eventually, you get enough ballistic photons through that you can map out an image.
Black Holes radiate at least Hawking Radiation and that radiation is supposed to reflect the information inside the Black Hole (and thus the information of some photon or other).
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
Actually, you'd want "does not conduct light". Something that transmits light, ie. originates it is different from something that merely acts as a conduit for already existent work
The truth shall always be free: Boris Floricic is Tron.
Yes, but so are most things if the light is bright enough or the detector sensitive enough.
Hmm... Didn't hurt that time.
Why is it so hot? Where am I going? What am I doing in this handbasket?
...years from now the technology could supplement or even replace traditional ultrasounds for baby imaging...
Ultrasound is for listening to sounds inside the body, such as the faint heartbeat of an unborn human. Imaging is accomplished by Sonogram.
Learning about brewing beer, by brewing beer.
To a google truck near you.
---- Booth was a patriot ----
It just so happens that your object here is only MOSTLY opaque. There's a big difference between mostly opaque and all opaque. Mostly opaque is slightly transparent. With all opaque, well, with all opaque there's usually only one thing you can do.
There's a difference between a photon every once in a while making it through and having enough to form an image. There are likely a lot of objects that you simply can't form a reasonable image through without using enough light to vaporize them.
It's called Ballistic light.
Eventually, you get enough ballistic photons through that you can map out an image.
And if you get the light strong enough, you resolve the opacity issue permanently, once the smoke clears.
but what about for thinner body parts, like hands or feet? A quick scanner for those might still be useful for replacing the need for xrays in some situations. It might be possible to have the light sent to include positioning information (for example, send a string instead of a single pulse) and correlate where it was seen with where it was sent from.
Go through their clothes and look for loose change?
I think that's what they're getting at.
It might work on hands, if they manage to reconstruct better images. If you try and do CT you're going to get a lot of artifacts from the bones though.
Exactly. And thank you so much for spelling "loose" correctly.
Doesn't this sound like Angel Light by Troy Hurtubise?
He said he sold the technology to France and everyone said he was crazy.
http://en.wikipedia.org/wiki/Troy_Hurtubise
No, the answer that human hands are translucent is not valid. Transparency and translucency imply the ability to somewhat discern what is behind the object.
The human hand does not allow that.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Pretty sure those assholes with the halogen headlights can see through several meters of solid rock pretty easily.
ISTR seeing footage of nuke tests where the flash is so bright pretty much everything in front becomes see through. Well, until it becomes vapour.
I want a list of atrocities done in your name - Recoil
In which case detecting them would tell you nothing about the object they passed through.
to think that US law enforcement still follows such an arcane principal as 'the law'
Actually in this sense, transmit is correct. As a physics grad I would certainly never use conduct for anything other than heat or electricity.
It's official. Most of you are morons.
Listening to sounds within the body is called auscultation.
Unless you're not listening to the noises themselves directly with a stethoscope (= auscultation), but listening to artificial reconstruction of noises produced by a machine.
When it's not practical to stick a stethoscope on it (like the heart of a *un*born baby), you can use Doppler effect to detect motion of the blood. This is then either shown with colours superimposed over the regular US-picture, or (after Fourrier transformation) on a frequency/time/intensity graph, or can be converted to audible frequencies and played on speakers. So yeah, you can "listen" to a heart using ultrasound (after some processing)., instead of sticking your ear on it (with a tube in between).
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Eventually, you get enough ballistic photons through that you can map out an image.
Physicists don't actually use terms like "opaque" very often. We are more likely to talk about material that is "highly absorbing" or "highly scattering". The human body contains lots of both.
One area where people have tried to apply this is in optical mamography: women's breasts are primarily fatty tissue that is highly scattering but very weakly absorbing, so you get a surprisingly large fraction of transmitted light. You have to do a huge amount of processing to deconvolve the scattering kernel, but when I worked in the area in the late '90's it was getting close to useful.
For people reading this who are female or who have wives or girlfreinds willing to go along, go into a dark room and hold a flashlight under your (partner's) breast. You'll be amazed by the amount of veinous structure and whatnot you can see. Squeeze the breast flat to get more detail. Insert joke here about how now you're in a dark room with a woman who has at least one breast exposed so you know what comes next...
Very athletic women with smaller breasts may not see much: the chest muscles are highly absorbing and any any photon that scatters into them is lost.
High-speed computation is making visible light a more useful medium of detection all the time, and the work described in TFA is an interesting step along the way.
Blasphemy is a human right. Blasphemophobia kills.
One expects the hidden 3D image reconstruction process to improve with multiple image capture sources at different angles:
This would imply that public surveillance cameras could be used in a "phased-array" configuration to provide data for hidden 3D image reconstruction, unless the photons are routed around the target volume using metamaterial fabrics.
Rapid adoption of hidden 3D image reconstruction technology could result in a commercial demand for metamaterial fabrics to provide pedestrians with relative privacy.
Harry Potter is a fashion setter with his father's invisibility cloak!
On the flip side, plastic surgeons may discover a new source of revenue: Commercial brand placements INSIDE of patients willing to sell ad space to sponsored plastic surgery services.
"Is that a gun in your pocket, or are you just happy to see me?" -- Mae West
DarkStarZumaBeachSurfinApocalypseWow
Not really. 1mm of just about any pure metal will not transmit any visible wavelengths till the light source is powerful enough to vaporize it.
If information wants to be free, why does my internet connection cost so much?
Tunneling is exponential in nature. If you get say 1 in 10 photons through a 1um think piece of material. Then you only get a 1 in 100 with 2um thick.
Now lets assume that the thickness is now 1mm. Now just one photon in 1x10^1000 gets through.
Now lets assume we have a red laser (700nm) that has a power output of the sun (3.846×10^26 W). Thats 1.3x10^45 photons per second. After waiting 1000 billion years, thats still only 4.5x10^55 photons. So none get through. You could use all the energy in the visible universe and still not a single photon gets through.
So in other words, there really is such a thing as truly opaque objects.
By the way. Much less than one in ten photons of 700nm light would get through most pure metals.
If information wants to be free, why does my internet connection cost so much?
Physicists don't actually use terms like "opaque" very often.
Yes we do.
If information wants to be free, why does my internet connection cost so much?
For many materials nothing ever effectively sneaks through at least in the visible region.
See my other post: http://tech.slashdot.org/comments.pl?sid=1811066&cid=33824520
If information wants to be free, why does my internet connection cost so much?
Dear Mr. DumbAss:
Thank you for your submission to Slashdot. Your stunning intellect and grasp of the English language makes it clear that you will go very far in your career. First, however, you need to look up the definition of opaque.
Idiot.
I did say that it would be a tiny fraction. The GP's definition of "opaque" was as follows:
In order to be truly opaque, two criteria must be met - every photon has to intersect a particle and for every such intersection, the particle has to be able to absorb the photon.
To me, that means that no photons can ever get through. Tunneling is a matter of probabilities, so one might get through, even if the odds are overwhelmingly against it.
"The state is that great fiction by which everyone tries to live at the expense of everyone else." - Bastiat
Not a single photon even after 100s of times the age of the universe is not a "tiny fraction" by any stretch of the imagination. Its none.
The *size* of a photon in the visible region is much larger than an atom. Blue light is in the 400nm range while an atoms radius is in the 0.1nm range. So a photon intersects many atoms.. often all at once.
If information wants to be free, why does my internet connection cost so much?
You are correct, but there you get into the muddy waters of whether you mean "enough to form an image" even in theory (ie: regardless of technological constraints) or whether it's constrained to some specific level of technology. The theory option then depends on how complete the theory is (you can only extrapolate so far before any theory breaks down) and whether there even is much in the way of theory when it comes to photons passing through matter.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
We can do a little better than theory. If you assume you can perfectly account for the scattering that happens when photons go through the subject you're basically left with a regular imaging problem. You've got some sort of detector, probably a CCD, and you've got photons hitting it. Your detector has such and such a quantum efficiency, and is subject to a certain amount of noise from various sources. Even if you postulate a perfect detector, at any temperature above absolute zero it's still going to be subject to thermal noise, which can be quantified, and that puts a hard limit on the minimum incoming light intensity you need.
Really, if you can find numbers for how much light is attenuated going through various materials (such as the skin, fat, bone, blood and brain in someone's head), assuming perfect reconstruction, the problem would come down to a simple exposure calculation that's more or less familiar to any photographer.
Ok, that works. We can remove the technology-dependence by describing it in the abstract. If you have a signal, you will have noise. We can then circumvent the question of the type of noise and the conditions surrounding it. If the noise conceals the signal, you will not be able to detect the signal.
Now, we can apply a bit of theory to this. If we treat each possible path from the object to the detector as a continuous-time analog channel and the presence/absence of some count of photons of a given frequency in a given interval of time as your digital data, you can use the Shannon-Hartley theorem to determine how often the apparent detection of a photon or block of photons will be correct.
Now it gets fun. If we can alter the frequency of the photons (say by using a different light source), the overall noise characteristics will be the same (the details may differ though). If the change is small enough, the diffraction will also be the same with the absorption differing. If you step up the frequency such that the absorption is the same, the diffraction will necessarily differ. You can therefore vary those two parameters independently. (This assumes that these two effects - fluorescence and diffraction - are the underlying cause of all other optical properties. You'd need to extend this if other fundamental properties exist.)
The idea here is that noise is random but information is not. Therefore for any set of data sets, the noise will be random on each but all the information must be related using well-defined rules. If you count the number of 1s and 0s (be they individual photons or blocks of photons) and place them into a bucket, some of that bucket will be noise and some will be data. If you have one data set, every possible division between noise and data could potentially be correct. As you increase the number of data sets, the number of possible divisions is reduced. You will never get to just one potentially valid division in the case of a totally swamped signal. If you can get down to a workable number of candidates, that may be sufficient. (For something as complex as the human body, there will always be multiple conditions that produce the same output - this just increases the range of potential causes.)
It's really a variant on the most basic form of error-correction (sending multiple copies), on the assumption that different types of signal will have errors in different places, allowing for more probable reconstruction.
Since the information is generated by the body the light passes through and not the source, adding any actual error-correcting codes to the data would be extremely hard if not impossible.
Oh, and no I don't think modern technology would be capable of producing a large enough set of data sets to be able to un-entangle the data from the noise.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
There's an easier way, which is commonly used: averaging.
You don't actually have to change the light frequency. Since the noise is random, simply imaging N times and averaging the result will give you a sqrt(N) improvement in SNR. Astrophotographers do this when they stack multiple images, and it's also used in MRI scans when you acquire the same image multiple times and average.
A continuous version just involves imaging longer. A standard digital camera will leave it's shutter (electronic or mechanical) open for a set period of time. To form an image of a given brightness, the CCD amplifier then has to amplify the recorded signal (signal + noise) by a certain amount (which is set by the ISO setting on the camera).
If you leave the shutter open longer, more photons accumulate in the CCD wells, and the amplifier has to amplify less (lower ISO). The CCD case is a little more complicated because some of the noise sources ALSO increase over a longer exposure, but some don't, so you get a less noisy image.
As in most imaging modalities, in this case the amount of averages (or the length of time you can image) is limited by what you're imaging. If it's not alive then go to town - scan it for days if necessary. If it is alive and your scan time is too long, it's probably going to move and ruin your picture. At some point it will hop off the table and want a hamburger.
Yeah, long baselines work. I was working on the assumption that at any given frequency there will be gaps in what you can know because of absorption so by changing frequency you can change what gaps there are. (Since you can't tell in a noisy signal whether the gap is due to noise swamping the data or there really being a gap.)
As for hamburgers, I think that anything moving would prefer something more substantial. :)
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
Is this another application of the Wiener filter?
http://en.wikipedia.org/wiki/Wiener_filter
The *size* of a photon in the visible region is much larger than an atom. Blue light is in the 400nm range while an atoms radius is in the 0.1nm range. So a photon intersects many atoms.. often all at once.
What you say about blue light (400nm) is its wavelength, not the size of the photon itself. Size and wavelength are separate properties. I could be wrong, of course (I am not particularly familiar with quantum mechanics and such).
It was a really good paper.
You are wrong. The wavelength is approximately the size. In fact you can treat light at this wavelength as a wave because its large compared to the atoms, and treat the material has a homogeneous solid. You get the same answer.
If you don't believe me, run the numbers yourself.
The conversion of doppler us to audio is a small part of one us modality though.
(Nowadays. Before miniaturization of digital processing, sound-only Doppler where the only small and hand-carried devices. I had met old Internal Medicine professors who only trust their antiquated sound-only portable device for some exams).
Even if you accept that you're actually listening to sounds, it's not true that "ultrasound is for listening to sounds inside the body" NOT imaging.
Yeah, you're right. I was just nit picking on the fact that you can use it to listen to heartbeats when a normal auscultation isn't practical.
But basically yes, to go back to your metaphor :
like listening to the sound a car makes by feeding the output of a radar gun into a speaker
Ultrasound is mostly used for imaging, just like radar guns' main usage is measuring cars' speed.
Yeah ! Car metaphors on /. !
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
There are very few opaque objects, people are translucent. A small percentage of photons manages to pass through without hitting anything which will stop them. Both intensity and frequency (color) deliver information.
BTW: there's some prior art on this, Dr Jerry Tiemann had proof of concept going prior to 1993, at GE's Corporate Research Center. GE Medical Systems declined to fund development. I believe he was using an algorithm developed by Dr Glen Row for transforming fan beam data to parallel beam, to produce more resolution with less computation. Jerry was disappointed that GE was putting resources into cold fusion at the time, instead of photon imaging. Yes, that's the father of Michael Tiemann, the Redhat CTO.