|
Home > Archive > Radiology > July 2005 > Sharpened X-ray Image
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Sharpened X-ray Image
|
|
|
|
| relkniw@bol.com.br 2005-03-23, 5:38 pm |
| I'm a physician, but not specialised in radiology. For me, sharpening
X-ray images, as shown in the website, is pointless. The "blurred"
images have been used for the last 100 years. As Bill wrote, there are
lots of artifacts. These can be not only worthless, but also dangerous,
leading to wrong diagnosis and treatment.
Tucano
| |
| tontoko 2005-03-23, 5:38 pm |
| I thought if the dimness of image correlates the shape and arrangement
of bones, it might be possible to synthesize stereoscopic image by
using it. The following images were synthesized by a software with
those 'sharpened" images.
http://www.geocities.com/q17320508/.../xray2/xray.htm
| |
| Tucano 2005-03-23, 5:38 pm |
| tontoko wrote:
> I thought if the dimness of image correlates the shape and arrangement
> of bones, it might be possible to synthesize stereoscopic image by
> using it. The following images were synthesized by a software with
> those 'sharpened" images.
> http://www.geocities.com/q17320508/.../xray2/xray.htm
>
I'd like to see the software working, but when I enter the website I
receive a msg asking me to turn Java on, but it's already on...
Anyway, I'm afraid that such stereo images produced by matematical
methods can be misleading for physicians. It's certainly more advisable
to obtain two X-ray incidences of the patients to have *real* spacial
information, than using algorithms.
Remember that X-ray are not only to see bones, but also soft tissues.
The last image appears to be a cerebral T1-weitghted MRI after
sharpening. The MRI scans are currently the best method to obtain highly
accurate images, with anatomic perfection. Here, even more than with
X-ray films, there is no need at all for sharpening, and, in fact, the
process of sharpening has made lots of valuable anatomical information
just disappear.
Think about it
Regards
Tucano
| |
|
| tontoko wrote:
> I thought if the dimness of image correlates the shape and arrangement
> of bones, it might be possible to synthesize stereoscopic image by
> using it. The following images were synthesized by a software with
> those 'sharpened" images.
> http://www.geocities.com/q17320508/.../xray2/xray.htm
>
If you got a blurred image, the only way to correct or improve is to
retake by:
1. Reducing motion
2. Reducing OID
3. Using smaller focal spot size
4. Using a detail screen (which needs more mAs)
5. Increasing SID
If density and contrast are concerned, you can manipulate the darkness
and brightness after the projection in CR and DR but not in films.
Regarding the idea of stereoscopic image, you don't get everything in
one image, so you are taking AP/PA, OBL, and LAT or R-OBL and L-OBL,
which are perpendicular to each other.
Some images such as scapulae, sternum, transthoracic humoral head, and
so on require blurred image of "ribs". Therefore, you need to instruct
the patient to breath slowly during the exposure. If the ribs are
blurred and unclear, the part of interest comes out sharply.
Your sharpening idea an work should be evaluated in extremity films.
Why don't you show us the films of finger, hand, wrist, forearm, elbow
joint, toe, foot, ankle joint, tib-fib, knee joint, and patella and
intercondylar fossa (skyline view) using your soft.
If you have, I would like to see facial bones, sinuses, and orbits
(blow-out if you have), and cranial bones too.
| |
| Repeating Rifle 2005-03-23, 5:38 pm |
| in article 1111298603.721698.275150@g14g2000cwa.googlegroups.com, tontoko at
gooday@hotvoice.com wrote on 3/19/05 10:03 PM:
> I thought if the dimness of image correlates the shape and arrangement
> of bones, it might be possible to synthesize stereoscopic image by
> using it. The following images were synthesized by a software with
> those 'sharpened" images.
> http://www.geocities.com/q17320508/.../xray2/xray.htm
>
I mangaged to view these images in stero by adjusting the convergence of my
eyes. I saw nothing usefully added. I do admit that one of my eyes has
uncorrectable por resolution.
Bill
| |
| Repeating Rifle 2005-03-23, 5:38 pm |
| in article BE61F7E0.358D4%salmonegg@sbcglobal.net, Repeating Rifle at
salmonegg@sbcglobal.net wrote on 3/19/05 3:38 PM:
> From what I see, the sharpening introduces various artifacts that are not
> present in the originals. The benefits of emphasizing edges look like they
> may be at the expense of showing things that are not there. The arrows and
> text were processed as well as the original images. It looks like those
> items greatly degraded in appearance. What point is there to the sharpening?
> What does a true expert have to say?
From what I see posted, I may be the partially sighted guy in the fable
about being in the land of the blind. :=)
No sharpening algorithm can give you more information than what is already
included in the unprocessed image. All it can do is to emphasize information
that is already there. False color, for example, may make images more
useful, but it does not actually add information. It only directs you to it.
From what I saw of the processed images, the artifacts arise from what is
called Gibbs phenomena well known in the application of fourier series. It
is possible to emphasize edges to some extent by emphasizing high spatial
frequencies and not introducing artifacts. This has to be done in a way so
as not to produce Gibbs phenomena. This is done by what is called
apodization. When you emphsize the high frequencies, it has to be done
smoothly.
Bill
| |
| tontoko 2005-03-23, 5:38 pm |
| OK. Although at the moment I don't have any other images "sharpened" by
that software, I'll prepare the images you've mentiened and announce
the results here later.
| |
| yucos_antalya3793@duskmail.com 2005-03-23, 5:38 pm |
|
gooday@hotvoice.com wrote:
> OK. Although at the moment I don't have any other images "sharpened" by
> that software, I'll prepare the images you've mentiened and announce
> the results here later.
Great
yucos_antalya3793@duskmail.com
| |
| Repeating Rifle 2005-03-24, 12:28 pm |
| in article BE61F7E0.358D4%salmonegg@sbcglobal.net, Repeating Rifle at
salmonegg@sbcglobal.net wrote on 3/19/05 3:38 PM:
> From what I see, the sharpening introduces various artifacts that are not
> present in the originals. The benefits of emphasizing edges look like they
> may be at the expense of showing things that are not there. The arrows and
> text were processed as well as the original images. It looks like those
> items greatly degraded in appearance. What point is there to the sharpening?
> What does a true expert have to say?
From what I see posted, I may be the partially sighted guy in the fable
about being in the land of the blind. :=)
No sharpening algorithm can give you more information than what is already
included in the unprocessed image. All it can do is to emphasize information
that is already there. False color, for example, may make images more
useful, but it does not actually add information. It only directs you to it.
From what I saw of the processed images, the artifacts arise from what is
called Gibbs phenomena well known in the application of fourier series. It
is possible to emphasize edges to some extent by emphasizing high spatial
frequencies and not introducing artifacts. This has to be done in a way so
as not to produce Gibbs phenomena. This is done by what is called
apodization. When you emphsize the high frequencies, it has to be done
smoothly.
Bill
| |
| Repeating Rifle 2005-03-24, 12:28 pm |
| in article 1111298603.721698.275150@g14g2000cwa.googlegroups.com, tontoko at
gooday@hotvoice.com wrote on 3/19/05 10:03 PM:
> I thought if the dimness of image correlates the shape and arrangement
> of bones, it might be possible to synthesize stereoscopic image by
> using it. The following images were synthesized by a software with
> those 'sharpened" images.
> http://www.geocities.com/q17320508/.../xray2/xray.htm
>
I mangaged to view these images in stero by adjusting the convergence of my
eyes. I saw nothing usefully added. I do admit that one of my eyes has
uncorrectable por resolution.
Bill
| |
| yucos_antalya3793@duskmail.com 2005-04-19, 10:44 am |
|
gooday@hotvoice.com wrote:
> OK. Although at the moment I don't have any other images "sharpened" by
> that software, I'll prepare the images you've mentiened and announce
> the results here later.
Great
yucos_antalya3793@duskmail.com
| |
| yucos_antalya3793@duskmail.com 2005-04-20, 5:56 pm |
|
gooday@hotvoice.com wrote:
> OK. Although at the moment I don't have any other images "sharpened" by
> that software, I'll prepare the images you've mentiened and announce
> the results here later.
Great
yucos_antalya3793@duskmail.com
| |
| John Lai 2005-07-12, 11:11 pm |
| Hello,
I just picked up this communication but cannot find the original thread.
What is the sharpening software you are referring to?
Regards,
John
<yucos_antalya3793@duskmail.com> wrote in message
news:d3t8gp$2nk$1@news2.netvision.net.il...
>
> gooday@hotvoice.com wrote:
>
> Great
>
> yucos_antalya3793@duskmail.com
>
|
| |
|
|