[Insight-users] possible underflow/overflow bug with itkGDCMImageIO

Walter Blume WBlume at stereotaxis.com
Mon Jun 2 11:53:53 EDT 2008


Thank you for the suggestion, it works.

Is there a simple way to "filter" the individual images to type short as they are read by the Series reader, so that I can avoid doubling memory usage?  I'm reading a set of 964 512x512 images into a volume.

Poking around the documentation I see reference to streaming, but looks like it is not supported for dicom readers.  Is that correct?

Thanks much for your help.

Sincerely
Walter

-----Original Message-----
From: Mathieu Malaterre [mailto:mathieu.malaterre at gmail.com]
Sent: Monday, June 02, 2008 7:29 AM
To: Walter Blume
Cc: insight-users at itk.org
Subject: Re: [Insight-users] possible underflow/overflow bug with itkGDCMImageIO

Hi Walter,

 You need to instantiate your image as type 'int'.

HTH
-Mathieu

On Wed, May 28, 2008 at 8:57 PM, Walter Blume <WBlume at stereotaxis.com> wrote:
> I am writing some code to compute DRRs, starting with the ITK example
> code in DigitallyReconstructedRadiograph1.cxx and
> DicomSeriesReadImageWrite2.cxx as examples.
>
> I have some DICOM images that are of type signed short.  They contain the
> value -32768 for some of the pixel values.   The rescale slope and intercept
> values in the dicom are 1 and -1024.   When I read these DICOM images the
> pixel values get scaled and transformed by the equation   buffer =
> pixel*slope + intercept.  This causes the buffer values to be
> represented as the positive number 31744 rather than being clipped to
> -32768, which then causes problems when computing the DRR, since I
> can't filter out the huge positive values.
>
> Is there a simple way to avoid this problem?
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users
>
>



--
Mathieu



More information about the Insight-users mailing list