Help:How to get raw CT values of a DICOM image using imageJ functions?

[razor]

Hello everyone!
Recently,i am doing a project and it seems that using the raw CT values(not gray scale value which range from 0 ~ 255) of a DICOM image as features can get a higher performance.But I don't kown use which function?
Hope for your reply!!

[razor]

Since it may be a little bit abstract,I will state it more clear.
What we want to do is to segment CT images(DICOM) and our temporary method is to use imageJ to convert a DICOM file to a 8-bit grayscale image(import image sequence -> set level to -600 and window to 1200 -> save as image sequence) , then use the converted image(.bmp) to do further work.
But there exists some "truncation" ------ this process convert the CT value in between (-600 - 1200 / 2) ~ (-600 + 1200 / 2) to 0 ~ 255 , and set others outside this range to 0 or 255 . Clearly , it isn't good for our algorithm.
So , my question is " How to get the raw CT values of CT images which can be directly read from DICOM files , is there some functions in imageJ or some other better method "

Thanks for you replies!!!

[Michael Schmid]

Hi razor,

you will never be able to convert an image with 12 bit resolution (as
typical for CT) to an 8-bit format like BMP without any losses.
Save it as tiff or zip (i.e., zip-compressed tiff) if you want to do
further analysis. This preserves the original data.

Many programs don't recognize image formats with more than 8 bits for
grayscale, so you may use an 8-bit format for importing into other
programs, e.g. for presentations, reports etc. But don't use the 8-bit
images for analysis!

Michael
__________________________________________________________________

On Sat, September 1, 2012 08:55, razor wrote:

> Since it may be a little bit abstract,I will state it more clear.
> What we want to do is to segment CT images(DICOM) and our temporary method
> is to use imageJ to convert a DICOM file to a 8-bit grayscale image(import
> image sequence -> set level to -600 and window to 1200 -> save as image
> sequence) , then use the converted image(.bmp) to do further work.
> But there exists some "truncation" ------ this process convert the CT
> value
> in between (-600 - 1200 / 2) ~ (-600 + 1200 / 2) to 0 ~ 255 , and set
> others
> outside this range to 0 or 255 . Clearly , it isn't good for our
> algorithm.
> So , my question is " How to get the raw CT values of CT images which can
> be
> directly read from DICOM files , is there some functions in imageJ or some
> other better method "
>
> Thanks for you replies!!!

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Volker Baecker]

Hello,
I never worked with DICOM images, but from your description I understand
that you use 8-bit grayscale. 8 bit means that you only have the range
from 0-255 available. The range you want to use is -1200 to 0 right? You
need to use 32-bit grayscale images to be able to represent the negative
values. If you use 32-bit grayscale instead of 8-bit and save in tiff
format instead of bmp you should be able to do what you want with ImageJ.

Volker
On 01/09/12 08:55, razor wrote:

> Since it may be a little bit abstract,I will state it more clear.
> What we want to do is to segment CT images(DICOM) and our temporary method
> is to use imageJ to convert a DICOM file to a 8-bit grayscale image(import
> image sequence -> set level to -600 and window to 1200 -> save as image
> sequence) , then use the converted image(.bmp) to do further work.
> But there exists some "truncation" ------ this process convert the CT value
> in between (-600 - 1200 / 2) ~ (-600 + 1200 / 2) to 0 ~ 255 , and set others
> outside this range to 0 or 255 . Clearly , it isn't good for our algorithm.
> So , my question is " How to get the raw CT values of CT images which can be
> directly read from DICOM files , is there some functions in imageJ or some
> other better method "
>
> Thanks for you replies!!!
>
>
>
> --
> View this message in context: http://imagej.1557.n6.nabble.com/Help-How-to-get-raw-CT-values-of-a-DICOM-image-using-imageJ-functions-tp4999912p4999913.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html