Conversion of the gray value (32bit float) back to 12-bit unsigned

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Conversion of the gray value (32bit float) back to 12-bit unsigned

Hansoo Chang
Hello,

I am in trouble in my project using ImageJ. I would appreciate it very
much if somebody could help me.

I imported a medical image (DICOM) that has been encoded with 12-bit
unsigned. After the import, the ImageJ shows the type of the image to be
32-bit.

I am measuring the gray values of the image in a ROI and obtaining the
average gray value inside the ROI.

The problem is that I need the absolute values of the original 12-bit
unsigned data for further data analysis because the flow speed of the
material is encoded into the original 12-bit unsigned data.

I have surveyed the internet for how I can convert the data obtained by
the ImageJ measurement back into the original 12-bit unsigned value in vain.

Would someone help me solve this problem?

Hansoo
--
---
Hansoo Chang
email: [hidden email]

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: Conversion of the gray value (32bit float) back to 12-bit unsigned

Stoyan Pavlov
Hello Hansoo,
If you import your images with the option "Scale when converting" in
Edit>Options unchecked , you should be able to keep the original Grey
values. Then when you convert to 16 bit, they still must be retained as
long as you don't do any permanent histogram (brightness and contrast)
manipulations. Just remember that upon restart "Scale when converting" is
always checked.
P.S. To my knowledge there is no 12 bit image format. We use either 16 bit
or 32 bit image formats to store 12 bit data.
Best regards,
Stoyan



На пт, 22 май 2020 г., 7:03 Hansoo Chang <[hidden email]> написа:

> Hello,
>
> I am in trouble in my project using ImageJ. I would appreciate it very
> much if somebody could help me.
>
> I imported a medical image (DICOM) that has been encoded with 12-bit
> unsigned. After the import, the ImageJ shows the type of the image to be
> 32-bit.
>
> I am measuring the gray values of the image in a ROI and obtaining the
> average gray value inside the ROI.
>
> The problem is that I need the absolute values of the original 12-bit
> unsigned data for further data analysis because the flow speed of the
> material is encoded into the original 12-bit unsigned data.
>
> I have surveyed the internet for how I can convert the data obtained by
> the ImageJ measurement back into the original 12-bit unsigned value in
> vain.
>
> Would someone help me solve this problem?
>
> Hansoo
> --
> ---
> Hansoo Chang
> email: [hidden email]
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: Conversion of the gray value (32bit float) back to 12-bit unsigned

Fred Damen
In reply to this post by Hansoo Chang
Greetings,

Short answer: https://imagej.net/DICOM_open

As you have not indicated the symptoms of your issue, I will assume you
suffered from what I did ages ago.  The voxel data is probably being
converted from 12 bit unsigned to what ImageJ thinks is 16bit unsigned
correctly, n.b., I deal with 16 signed voxels and do not have 12 bit to
test this. (If your data is 12-bit information stored in 16-bit chunks,
your ok.  If its 12bit information stored in 12 bit chunks then you still
might have an issue - there are tools that read DICOM and place in easy to
read binary format.)  After ImageJ converted to 16 signed, it is converted
to 16 bit unsigned by adding 32768(why????). If the DICOM tags for scaling
exist the voxel data can be accessed with getPixelValue to get the correct
scaled value.  Trouble for me is that they don't always exist for my
datasets.

Enjoy,

Fred

If the above is confusing, open the DICOM dataset and do a Image>Show
Info... and send; Make sure you delete any patient info first though.

On Thu, May 21, 2020 10:51 pm, Hansoo Chang wrote:

> Hello,
>
> I am in trouble in my project using ImageJ. I would appreciate it very
> much if somebody could help me.
>
> I imported a medical image (DICOM) that has been encoded with 12-bit
> unsigned. After the import, the ImageJ shows the type of the image to be
> 32-bit.
>
> I am measuring the gray values of the image in a ROI and obtaining the
> average gray value inside the ROI.
>
> The problem is that I need the absolute values of the original 12-bit
> unsigned data for further data analysis because the flow speed of the
> material is encoded into the original 12-bit unsigned data.
>
> I have surveyed the internet for how I can convert the data obtained by
> the ImageJ measurement back into the original 12-bit unsigned value in
> vain.
>
> Would someone help me solve this problem?
>
> Hansoo
> --
> ---
> Hansoo Chang
> email: [hidden email]
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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