Rescaling Dicom images - DICOM

This is a discussion on Rescaling Dicom images - DICOM ; Hello! I have encoutered a strange problem a i have tried to rescale Dicom images: after reading dicom series of CT files, the orignal images have a modality rescale slope of 1.0 and a modality rescale intercept of -1024. The ...

+ Reply to Thread
Results 1 to 4 of 4

Thread: Rescaling Dicom images

  1. Rescaling Dicom images

    Hello!

    I have encoutered a strange problem a i have tried to rescale Dicom
    images:

    after reading dicom series of CT files, the orignal images have a
    modality rescale slope of 1.0 and a modality rescale intercept of
    -1024. The pixel range is -1024 up to ~1200 (the have all different max
    value).

    Now after writing the series again to dicom slices, there are some
    images that are totaly white.

    Has anybody an idea what could cause these strange resaling image
    values?

    Thanks in advance


  2. Re: Rescaling Dicom images


    mlouizi@gmail.com wrote:
    > Hello!
    >
    > I have encoutered a strange problem a i have tried to rescale Dicom
    > images:
    >
    > after reading dicom series of CT files, the orignal images have a
    > modality rescale slope of 1.0 and a modality rescale intercept of
    > -1024. The pixel range is -1024 up to ~1200 (the have all different max
    > value).
    >
    > Now after writing the series again to dicom slices, there are some
    > images that are totaly white.
    >
    > Has anybody an idea what could cause these strange resaling image
    > values?


    What toolkit did you used for the conversion. Are you sure you can
    trust the conversion ? Also as a shot in the dark, if your image is
    declared as unsigned short and you try to apply a shift of -1024 on
    unsigned pixel type ... you might have some surprise

    Be sure also ro remove any rescale/intercept (or set them to 1 & 0) in
    the image you have rewritten.

    After that, I don't know. I would need to have a look at the original
    data and the rewritten one...

    HTH
    Mathieu


  3. Re: Rescaling Dicom images


    Mathieu Malaterre schrieb:

    > mlouizi@gmail.com wrote:
    > > Hello!
    > >
    > > I have encoutered a strange problem a i have tried to rescale Dicom
    > > images:
    > >
    > > after reading dicom series of CT files, the orignal images have a
    > > modality rescale slope of 1.0 and a modality rescale intercept of
    > > -1024. The pixel range is -1024 up to ~1200 (the have all different max
    > > value).
    > >
    > > Now after writing the series again to dicom slices, there are some
    > > images that are totaly white.
    > >
    > > Has anybody an idea what could cause these strange resaling image
    > > values?

    >
    > What toolkit did you used for the conversion. Are you sure you can
    > trust the conversion ? Also as a shot in the dark, if your image is
    > declared as unsigned short and you try to apply a shift of -1024 on
    > unsigned pixel type ... you might have some surprise
    >
    > Be sure also ro remove any rescale/intercept (or set them to 1 & 0) in
    > the image you have rewritten.
    >
    > After that, I don't know. I would need to have a look at the original
    > data and the rewritten one...
    >
    > HTH
    > Mathieu


    Hello Mathieu,

    sorry but i think i have not explained my problems really good.

    I have DICOM series of CT. In CT images the paramters rescaleIntercept
    and rescaleSlope are generally set to respectively -1024 and 1.0. In my
    case there are some images that have a slope value of 0.5.
    This results in bright images/slices when showing the 3D volume.

    Thus the deal is to rescale these "wrong" images to the right values
    area.

    Do you have any idea to resolve this problem (using ITK library!!)?

    thanks


  4. Re: Rescaling Dicom images


    simoloui wrote:
    > Hello Mathieu,
    >
    > sorry but i think i have not explained my problems really good.
    >
    > I have DICOM series of CT. In CT images the paramters rescaleIntercept
    > and rescaleSlope are generally set to respectively -1024 and 1.0. In my
    > case there are some images that have a slope value of 0.5.
    > This results in bright images/slices when showing the 3D volume.
    >
    > Thus the deal is to rescale these "wrong" images to the right values
    > area.
    >
    > Do you have any idea to resolve this problem (using ITK library!!)?


    Hi Mohamed,

    I'll answer you on the ITK mailing list instead, since this newsgroup
    is more general.

    Mathieu


+ Reply to Thread