Hello all,
I am having the following issue.
In general I use Amide (a free software package) to view my DICOM files.
It is pretty good to study dynamic PET-images and okay in drawing ROIs.
My interest lies in the evolution of signal intensity over the course of time for ROIs I draw on the images.
Because of some scan software issues I need to correct all my images with a time dependent factor (something to do with decay correction). I have written a script in ImageJ which can do this fine. But now I want to export this image sequence again as slices with a DICOM header/information that recognises that the slices have xyzt coordinates.
I only manage to get a DICOM file in which the t is not condisered and all my slices (x,y) are arranged in a row (z).
In my case I have for instance 1260 images which consist of 200 x , 200 y pixels. The depth of the 3D scan is 63 slices and an 3D image is generated each 5 minutes for 100 minutes so I have 20 t-point yielding 1260 x,y images.
In a nutshell;
- I import dynamic 3D images in ImageJ
- I correct them with a ImageJ script
- I want to export them with time information in xyzt format.
Hopefully anybody has some inspiration to resolve these issues.
Thanks a lot.
Kind regards,
Michiel
Image5D is an ImageJ plugin that handles up to 5 dimensions. It might do the trick. You might also need to save the image as a TIFF file, and then convert it to DICOM using ImageMagick.
Hello Michiel,
I have a problem that I think you have solved. I have dynamic PET-images and I want to quantify some parameters using ROIs in imageJ/Fiji but when I import my images in imageJ it loses its quantitative information about Time Activity Curves. Could you tell me how did you solve that problem writing a script? I don`t understand what is imageJ doing? Thanks a lot.
Hi Sarahi,
I would gladly try to help you.
Eventually I have used another solution than the ImageJ/Fiji solution.
Maybe if you can explain the precies context of what you want to measure or determine I can give you my opinion on how this would be best tackles.
Hello Michiel,
I am trying to get time activity curves (evolution of activity concentration over the course of time) from brain regions of dynamic PET-images, I can do it one by one drawing ROIs on the images directly in the workstation of the PET/CT system (using the comercial software syngo) but I want to do it faster using ImageJ or Matlab but when I import my DICOM files on imageJ and I mesure the gray intensity evolution over the course of time the data does not correspond with the expected curves.
I did an experiment with a dynamic images of F-18 phantom to mesure the decay rate and with imageJ I did not get the exponential decay curve expected.
I think that I need to rescale the images but I don’t know how to do it writing a script in imageJ. Is it a linear transformation? Do I need to do a calibration curve to correlate the activity concentration I mesure using syngo with the gray intensity I mesure using ImageJ?
Thank you for answering,
Kind regards,
In my experience, most software do the correction automatically.
In that case you need to select the correct radionucleide and then in case of a phantom you should be getting a flat line.
The scanner I used had a software bug in which sometimes the decay-correction was doubled and thus overcorrected.
The solution I found best was correction the image matrices in matlab.
I wrote a script I would gladly share with you, but I am nowhere near an experienced programmer nor a very experienced medical image analyser.
So the script is not that clean and it will probably need some tweaking in order to work for you as well as it did for me.
But once again for clarity; Are you mainly interested in correcting your images (in other words, are you sure your images have an error) or are you mainly interested in some kind of automatisation to select a ROI. In the latter case i prefer using Amide, which is a free software. It is absolutely not perfect but in many cases it does do the trick.
If you would like me to take a look at some of your images to check whether they look erroneous feel free to share some images.
Kind regards.
Dear Michiel,
I am really sorry for the delayed response, I was disconnected from this network for long period of time. I will try to use this network more, is very useful.
You were right, my images were not wrong, the correction is made automatically as you said.
Thank you for your kind answer.
Best regards,
- Badges
- Science topic
- Similar topics
- Journal Impact Factor