Follow Slashdot stories on Twitter

 


Forgot your password?
Close
typodupeerror
×
Medicine Technology

Single Pixel Camera Takes Images Through Breast Tissue 81

Posted by timothy from the one-at-a-time-folks dept.
KentuckyFC writes Single pixel cameras are currently turning photography on its head. They work by recording lots of exposures of a scene through a randomising media such as frosted glass. Although seemingly random, these exposures are correlated because the light all comes from the same scene. So its possible to number crunch the image data looking for this correlation and then use it to reassemble the original image. Physicists have been using this technique, called ghost imaging, for several years to make high resolution images, 3D photos and even 3D movies. Now one group has replaced the randomising medium with breast tissue from a chicken. They've then used the single pixel technique to take clear pictures of an object hidden inside the breast tissue. The potential for medical imaging is clear. Curiously, this technique has a long history dating back to the 19th century when Victorian doctors would look for testicular cancer by holding a candle behind the scrotum and looking for suspicious shadows. The new technique should be more comfortable.
This discussion has been archived. No new comments can be posted.

Single Pixel Camera Takes Images Through Breast Tissue More

Single Pixel Camera Takes Images Through Breast Tissue

Comments Filter:

  • Strange KFC advertising (Score:1)

    by Anonymous Coward

    You aren't even mentioning if the chicken breast is deep fried.

  • more comfortable, but less fun (Score:5, Funny)

    by Trepidity ( 597 ) <[delirium-slashdot] [at] [hackish.org]> on Friday November 28, 2014 @07:34PM (#48482265)

    Some people pay good money to have a candle held behind their scrotum.

  • Not Human Breasts...Doh (Score:3, Insightful)

    by cb88 ( 1410145 ) on Friday November 28, 2014 @07:34PM (#48482267)
    Its chicken breast tissue.... thats just blatant click bait. However you want to look at it.

  • Not a camera (Score:5, Interesting)

    by graphius ( 907855 ) on Friday November 28, 2014 @07:39PM (#48482285) Homepage

    Not really a single pixel camera, more of a single pixel light absorption meter taken over an area...

    • Re:Not a camera (Score:5, Informative)

      by donaggie03 ( 769758 ) <d_osmeyer @ h o tmail.com> on Friday November 28, 2014 @08:49PM (#48482555)

      Not really a single pixel camera, more of a single pixel light absorption meter taken over an area...

      What is a camera if not a glorified light absorption meter?

      • Re: (Score:1, Offtopic)

        by graphius ( 907855 )

        In a way you are right, but there is a lot of difference between a camera and a light meter.

        Speaking as a photographer...

    • Re: (Score:3)

      by fche ( 36607 )

      The trick is that the light source varies with different samples. What this apparatus appears to be computing is a dot product (overall image intensity) with a series of 2D wavelets. Then inverse-transform the coefficients to a 2D image.

  • Victorian doctors would look for testicular cancer by holding a candle behind the scrotum

    I hope that they actually held the candle in front of the scrotum and looked from behind.

  • you had me at c.
  • A 40 megaray light sensor may be interesting in such applications?
    http://www.lytro.com/ [lytro.com]
    Recording more than just intensity per pixel...

  • Maybe it's because I've been drinking since 10am, but I just cannot wrap my head around how these single-pixel cameras work.

    Any nice person out there feel like explaining it so a stupid person can understand?

    Bonus points if you explain why a chicken breast was involved. Seriously, maybe I have brain damage because I've read the summary three times and it might as well be written in Middle Egyptian for all I'm getting.

    • I just cannot wrap my head around how these single-pixel cameras work.

      Well, the camera only has a single pixel . . . but . . . they bounce the light off an "array of digital micro-mirrors". And they re-arrange the "array of digital micro-mirrors" and take a new shot 500 times. So just think of them as just taking 500 random pixels samples from the whole picture, and number crunching it, to smooth it out. Note, this is an over-simplification.

      Bonus points if you explain why a chicken breast was involved.

      The summary mentioned the "candle in the crotch" crew, but TFA also mentioned that the same method was used to detect breast cancer by

      • The candle was only mentioned in the summary, the article refers to a lamp being used to backlight tissue to look for telltale abnormalities in the scattered light, which seems like a light source far better suited to the task.

        So while you may well be right about a predilection for candle-related shenanigans being involved, I think it's in the minds of the submitter or editors, not the medical researchers or Victorian doctors. Or perhaps it's only that their preconceptions of the Victorian era skewed their

    • Re: (Score:2)

      by skids ( 119237 )

      Shoot a randomly speckled light pattern through a splitter. Put one copy of the pattern into a multipixel camera, then multiply what the multpixel camera saw with what the single pixel receiver saw after the second copy of the pattern bounced off or through the target. Rinse repeat, sum that array of pixels over lots of iterations. Basically that tweaked with additional statistics and physics for efficiency/accuracy. Perhaps eliminate the multipixel camera if you can find some other way to know the speck

      • Re: (Score:2)

        by jfengel ( 409917 )

        I got the impression that the idea was to use the breast itself as the scattering medium, in an attempt to focus on something inside the breast, namely a cancer. The chicken breast was used to simulate breast material, which would probably be pretty clever if it weren't so confusing.

        Chicken breast is very different from mammalian mammary tissue, but it was probably the cheapest source of meat. It may also help that it's relatively uniform, to simplify things, while simultaneously being sufficiently random a

        • I don't think it was even used to simulate breast tissue specifically, just generic "living tissue", as opposed to the frosted glass they started with. The only place "breast" is mentioned in the article without being proceeded by "chicken" is a single line commenting that the Victorian era lamp-behind-the-testicles technique was also used to detect breast cancer.

    • Re:Single-pixel what? (Score:5, Informative)

      by Arkh89 ( 2870391 ) on Friday November 28, 2014 @10:01PM (#48482833)

      Ok, let's say that you want to build a 1 "mega-pixel" camera (1000x1000 pixels, for instance). You have the optics but not the sensor array. Instead, you only have a single photo-diode... which is basically a single pixel.

      First approach : you decide to scan the image plane with this photo-diode, trading spatial resolution for time. You move the photo-diode to where the first pixel in the top-left corner of the sensor should be, integrate (collect the photons) for some time, then move to the second pixel position. After making 1 million of such movements/integrations, you have fully sampled the image plane and have a complete 1 "mega-pixel" image.
      Problem : this is slow as hell, you need to move the photo-diode up to some accuracy, etc.

      Second approach : instead of moving the photo-diode you will modulate the incoming signal (photons) and integrate everything to this detector. You take a small video projector and open it to find a component called a DMD which is an array of controllable bistable micro-mirrors. Basically, displaying an image on the video projector is turning this surface as a transmissive gray-scale pattern (note that it is not actually transmitting light, just reflecting). You put it in the image plane (at the position of the sensor array) and you use a lens to focus all of the light coming out of the DMD surface onto the photo-diode.
      Now, instead of scanning, you just have to display a pattern consisting of a "black" frame (fully "blocking") except only one "white" pixel ("transparent") and integrate as usual. As you know which patterns was used for each integration and can, as previously, rebuild the image.

      Second approach, first improvement : instead of lighting pixel per pixel you can use specific patterns. The basic idea is to integrate photons coming from multiple pixels at the same time and reconstruct with a specific algorithm. The idea is to express the problem as a linear equation A x = y where x is the input image, A is the measurement operator = a matrix representing the system and y is the measured vector. In the previous case, you were measuring pixel per pixel which is equivalent as modelling A as the identity matrix (ones on the main diagonal, zeros everywhere else and so y = x). Imagine now that you use another matrix / another way to combine multiple pixels, such that each row of A is pattern you have to display on the DMD and the matrix row is still square and full-rank (a well defined system). In the end you can still reconstruct x from y with A' y = x (where A' is the inverse of A) and get back your image.
      Why would you do this? Well, instead of getting a bunch of photon from a tiny opening you will be measuring many more photons which is a good thing as our real-world detector is noisy. You will thus increase the signal to noise ratio.

      Second approach, third improvement : the main problem of the previous system is that, to obtain a 1 mega-pixel image, you still need to do 1 million projections/measurements which is a lot, and makes the whole process slow. But, you know for a fact that images are compressible signals (JPEG is a proof of that) which means that you can represent any 1 mega-pixel image signal into a much smaller vector size. This is because natural images are not random structures and possess some level of coherency = redundancy between pixels. So instead of making as many projection as they are pixels (a square matrix), you will do less, say by a factor between 4 and 10. The matrix A becomes rectangular and you have to use a more complex reconstruction algorithm (non linear, or based on a convex optimization system) which takes into account prior knowledge you would have of natural images (think of it as external constraints that will help you make the system sufficiently well behaved).

      This is basically how single-pixel cameras work (with compressive sensing)...

      I'll pass for the bonus point.

      • You move the photo-diode to where the first pixel in the top-left corner of the sensor should be

        OK, so the sensor is moving. That helps. It's like a dot matrix printer in reverse? I got hung up on the word "camera", I guess.

        • No, the sensor isn't moving, except in case 1 in which the principle is illustrated. What it sounds like to me is that instead of moving the sensor, you project combinations of patterns onto the lens such that it cancels out every area except the one you want to sample. You subsample the image area and then interpolate the contents.

          • Re: (Score:1)

            by Anonymous Coward

            Not exactly. Rather, the DMD (an array of MEMS mirrors that can allow or block light from reaching the image plane at a single point) mentioned in the earlier post is configured to one of a set of known pseudorandom patterns for each captured frame. This means that each frame captured by the single-pixel sensor is the total of all the photons that reflect off of each activated mirror and reach the image plane.

            You repeat this process a large number of times, each time with a different pattern in the DMD.

            • You know, this Thanksgiving weekend, I'm grateful for the number of very smart yeggs what hang out here at Slashdot. It's honestly inspiring.

              I ask a question and I actually got patient, thorough explanations, on the Internet. You don't find that over at the Twitter or 8chan.

      • Re: (Score:2)

        by jfengel ( 409917 )

        Thank you for that incredibly lucid explanation. (Sorry, no mod points today, but you were at 5 anyway.)

    • Re: (Score:2)

      by Mandrel ( 765308 )

      Bonus points if you explain why a chicken breast was involved.

      Chicken breast, like most tissue, is translucent, and they were trying to demonstrate the ability to take a picture of a target hidden under several mm of such material, at the same time demonstrating the technique's applicability to diagnostic imaging.

    • As I read it, chicken breast was simply used as an organic scattering medium. Initial experiments were with frosted glass, but if the eventual goal is a new medical imaging technology then it really needs to be regularly tested with the kinds of tissue it would be working with in the field. Presumably they'll eventually start working with pork steaks, organ meats, Rocky Mountain Oysters, etc., but chicken breast is nice, relatively uniform, muscle tissue to get started with, and it possesses a very differ

  • Dual Photography (Score:5, Interesting)

    by Required Snark ( 1702878 ) on Friday November 28, 2014 @08:49PM (#48482549)
    Researchers from Stanford demonstrated in 2005 how to generate an image of a scene from the point of view of the light source instead of the camera. It's called dual photography, [stanford.edu] and has some similarities to the single pixel technique.

    We present a novel photographic technique called dual photography, which exploits Helmholtz reciprocity to interchange the lights and cameras in a scene. With a video projector providing structured illumination, reciprocity permits us to generate pictures from the viewpoint of the projector, even though no camera was present at that location. The technique is completely image-based, requiring no knowledge of scene geometry or surface properties, and by its nature automatically includes all transport paths, including shadows, interreflections and caustics. In its simplest form, the technique can be used to take photographs without a camera; we demonstrate this by capturing a photograph using a projector and a photo-resistor. If the photo-resistor is replaced by a camera, we can produce a 4D dataset that allows for relighting with 2D incident illumination.

    It exploits Helmholtz reciprocity to swap the camera view with the light view. If light is modeled as rays/photons, the path between the light source and a camera pixel is the same going from the light to the pixel, or the pixel to the light. Hence reciprocity.

  • Probably just a scanning lightsource instead of those patterns would work even better.

  • "Chicken chest! Chicken chest!" -- Archie Bunker

  • You know, on chickens, breast tissue is usually called breast meat. I've only seen it called breast tissue on mammals, and then usually only on females.

  • I have a feeling this technology will be first used to examine breast tissue hidden behind objects.

Slashdot Top Deals

Don't tell me how hard you work. Tell me how much you get done. -- James J. Ling

Close