View Full Version : How to scan x-rays revisited...

Virginia Ferguson
07-20-2000, 05:45 AM
Hello All:

Thank you for all of the many responses to my query regarding scanning
x-rays. This email summarized the responses for those who were interested
in knowing the outcome of this quandry.

In review, I wanted to scan x-rays of small animals (mice or rats) and
analyze them for morphology and bone density. Standard, commercially
available scanners do not have the power to scan x-rays and maintain
resolution and coloring. I tried to place several different lightsources
over the scanner with little luck - the lightsource would change over time
(i.e. the scanner could detect temperature changes in the light over time -
which would make the bony portion of the x-rays appear brighter over time).
I wanted to analyze these x-rays for bone density (via the color intensity
of the bony areas) and for shape (i.e. cortical thickness). The changes in
the light source over time - as well as the uneven distribution of light
over the surface of the x-ray (resulting in lighter and darker regions) made
this solution unacceptable for use in small animals (although it may work
for larger animals and humans).

Here is a summary of the responses and my comments on each:

1) Try having a mirror cut to match the dimensions of your scanner. Place
x-ray on the scanner and then place the mirror over it. The quality isn't
absolute top notch, but it ain't bad and is certainly cheaper than buying a
scanner dedicated for x-rays! This isn't difficult, any glass shop can do
it no problem and inexpensively.

*** We tried using a large mirror with several different HP scanners (HP
IIC, HP 5C, and an HP 5300C). This did not work well. In fact, several
responders told me that films are best scanned using transmitted light,
rather than reflected light.

2) Use the x-ray as a photo negative and print a contact print -
ie place the x-ray on some 10x8 photo paper, expose it to
white light, and develop the print as you would any photo. This print can
then be scanned in by a conventional scanner.

*** We tried this through our Universities publishing department - it
resulted in a high degree of variability on the contact prints when
processing the same x-ray more than once. Also, it does not pick up areas
where bone is scarce (that can be seen on the x-ray) such as the ulna,
radius, and hand bones of a mouse.

3) Take pictures with a high quality digital camera when they are on a
light board, etc.

*** A light board provides an uneven lightsource. If using the images for
the purpose of publishing an x-ray, this works fine. If you want to analyze
for intensity, this is not acceptable, in my opinion, as - once again - the
gradations in the light source cause some areas of the image to appear
artifically lighter.

4) Lay a lightbox on top of the x-ray on the scanner.

*** Same as #3. Also, light boxes with flourescent lamps can form
diffraction patterns on the scanned image.

5) One lab bought a $60 000 x-ray scanner that does the
job - results are at: http://www.health.latrobe.edu.au/Schools/POD/home.htm

*** These are expensive medical image scanners that will probably do a much
better job than commercially available scanners + transparency adapters. If
you need very high resolution and can afford it, this may be a good
solution. There are many schools that have such equipment in radiology

6) Most people who responded to my query suggested using a transparency
adapter. This is a special cover that works with your scanner to provide a
transmitted light source that tracks the scanning bar. Companies (and
scanners) recommended by responders were:

Hewlett Packard at
UMAX Astra 1220S scanner with ATA-3 transparency adaptor (www.umax.com) (up
to 8"x10").
UMAX Powerlook III Flatbed scanner.
Umax Mirage IIse professional (up to 12"x17")
Lumiscan 150 the x-ray film scanner

*** It turns out that I was able to obtain a scanner and transparency
adapter for free to test these suggestions. It is an HP Scanjet 5300C with
a transparency adapter. Thus far, it is working very well - excepting the
small area that the transparency adapter covers (5”x5”). The scanned images
appear to be consistently “lit” - resulting in no lighter/darker spots on
the image. And I can “normalize” all of my images by using a phantom - I
made an epoxy disk containing small sections of cortical bone with
thicknesses of 0.25, 0.5,...mm which I scan in on every image and use to set
a standard intensity when I analyze my images in Photoshop. Thus far, this
seems to be the most cost effective result that yields good results.

If you do not have the resources to buy such a system (much less than
$1000), then you can pay a professional service to do this for you and some
universities have publishing departments that do this as well.

Good luck and happy scanning,

Virginia Ferguson, M.S.

Orthopaedics Research Group
BioServe Space Technologies
Campus Box 429
University of Colorado at Boulder
Boulder, CO 80309

Phone: 303.735.2494
Fax: 303.492.8883

To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l