<br><font size=2 face="sans-serif">Hi Remi,</font>
<br>
<br><font size=2 face="sans-serif">CT values are measured in HU and can
be negative (water is 0, so air is negative). When stored in DICOM format,
these values are encoded to another pixel format (usually unsigned integers),
called "stored values", by applying a linear transformation.
Each CT DICOM image specifies the transformation from the stored values
to HU by providing Rescale Slope (tag 0028,1053) and Rescale Intercept
(tag 0028,1052):</font>
<br>
<br><font size=2 face="sans-serif">HU = m*SV+b</font>
<br>
<br><font size=2 face="sans-serif">Here m is the slope and b is the intercept
(SV stands hare for a stored value).</font>
<br>
<br><font size=2 face="sans-serif">Now when you have the values in HU,
you need to display them. For this purpose window center and window width
tags are used.</font>
<br><font size=2 face="sans-serif">Window Center (0028,1050) and Window
Width (0028,1051) specify a linear ramp that converts from HU values to
the grey levels to be displayed. Usually the user will adjust these values
interactively.</font>
<br>
<br><font size=2 face="sans-serif">Generally, there are few more tags specifying
different pixel formats in DICOM (signed/unsigned integer, number of relevant
bits per pixel, where the relevant bits are located, etc.), but in the
common case the situation is simple.</font>
<br>
<br><font size=2 face="sans-serif">Hope this helped,</font>
<br><font size=2 face="sans-serif"><br>
Mark Rabotnikov<br>
EBW team<br>
CT Engineering<br>
Philips Medical Systems Technologies LTD<br>
Phone: +972-4-8310646<br>
</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=33%>
<br>
<br>
<br>
<br>
<br><font size=1 face="sans-serif"><b>Remi Vieux <revieux@yahoo.fr></b>
</font>
<p><font size=1 face="sans-serif">Sent by:</font>
<br><font size=1 face="sans-serif">insight-users-bounces+mark.rabotnikov=philips.com@itk.org</font>
<p><font size=1 face="sans-serif">13/03/2006 06:07</font>
<td width=66%>
<table width=100%>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">To</font></div>
<td><font size=1 face="sans-serif">Itk mailing list <insight-users@itk.org></font>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">cc</font></div>
<td>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">Subject</font></div>
<td><font size=1 face="sans-serif">[Insight-users] Question about DICOM
and CT scanner</font>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">Classification</font></div>
<td></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<div align=right>
<br></div></table>
<br>
<br>
<br><font size=2><tt>Hi everyone,<br>
<br>
My question is not about ITK but I thought somebody here could help me<br>
with that: it's about CT scanner and DICOM image file format.<br>
<br>
So, basically, what I learned during my lectures on medical imaging is<br>
that a CT scan acquire its data, in Hounsfield Units, which may vary<br>
depending on the actual scanner you are using. Then, depending on what<br>
contrast you want to emphasize, you can choose a combination of window<br>
width/window level, and save this view into a DICOM file, hence mapping<br>
the H.U. into grey levels from 0 to 4095.<br>
<br>
Now, I'm doing a project for my university dealing with CT images in<br>
DICOM format. First thing I realised in the exams I have in my hand is<br>
that the data I'm retrieving are not grey levels from 0 to 4095 but from<br>
-1000 to 3095. Ok, not much harm done there, it is still 4096 grey<br>
levels. But then, answering to a previous question of mine in this<br>
mailing list, I've been told that I could not assume that it is always<br>
the case (or so I understood).<br>
<br>
So my question is: what exactly is saved in a DICOM file during a CT<br>
scan exam? Is it the 4096 grey levels representing a "windowed"
view of<br>
the exam as I was teached? Is it the H.U. as I thought it could be for
a<br>
moment? Is it both? None? Whatever..?<br>
<br>
That was quite a long speech for a short question, but, pardon me, I'm<br>
kind of lost here... If anybody could help me out, answering it or<br>
simply pointing me some reference in the litterature, I would deeply<br>
appreciate it!<br>
<br>
Best Regards,<br>
<br>
Remi.<br>
-- <br>
Remi Vieux <revieux@yahoo.fr><br>
<br>
<br>
PS: By the way, if somebody knows a good algorithm to perform a<br>
segmentation of the teeth in a CT exam, it would also be appreciated,<br>
but that's just because I'm stuck :-p<br>
<br>
<br>
<br>
<br>
<br>
<br>
___________________________________________________________________________
<br>
Nouveau : téléphonez moins cher avec Yahoo! Messenger ! Découvez les
tarifs exceptionnels pour appeler la France et l'international.<br>
Téléchargez sur http://fr.messenger.yahoo.com<br>
_______________________________________________<br>
Insight-users mailing list<br>
Insight-users@itk.org<br>
http://www.itk.org/mailman/listinfo/insight-users<br>
</tt></font>
<br>