Kodak K9500 CBCT

Home > Forum > Image Processing
3 replies [Last post]
sk26
Offline
Last seen: 1 year 33 weeks ago
Joined: 07/20/2009

I am working extensively with CBCT scans from the Kodak K9500, and I am finding these images to be almost too noisy to use for modeling.  If anyone else is working with this scanner please let me know, I'd like to compare images to see if it is just the machine at our hospital or it is typical of this scanner.  The example images provided by Kodak are far clearer than those we are getting in-house; although I have sent some images to Kodak, they do not seem to think the image quality is that bad.
Any thoughts?
 

0
LarsBonitz
Offline
Last seen: 35 weeks 4 days ago
Joined: 06/30/2009
CBCT
In our department we have Sirona/ Galileos and Newtom 3G. Pictures from both devices in comparison with CT include a high noise content.  The non optimal picture quality based on less radiation dose and force a non optimal discriminationgray of grey values. For the diagnosis and treatment, this low radiation  is sufficient. For construction of models and prosthesis the quality is not enough. Further development is needed.
sk26
Offline
Last seen: 1 year 33 weeks ago
Joined: 07/20/2009
K9500 vs NEWTOM 3G

I have mode direct comparisons of several CBCT devices to a Philips MDCT device (just about to submit a paper on that to OOOE).  Yes, the image quality is poor in comparison, but I have found the Newtom 3G and the Iluma images to have adequate quality for modeling, albeit with much hand segmentation.
The K9500 has a terrible problem with radial "ringing" artifacts, similar to Gibbs Artifacts, but manifested in a polar fashion in the FOV.  Rings and radial lines with high HU values, ruin much of the image.  I have not been able to filter these out using the filters supplied with MIMICS, I need something more sophisticated (or Kodak needs to reconsider its reconstruction approach).
I have taken this issue up with Kodak/Carestream, but, to date, they have not been forthcoming with an answer.  So I must assume that it is not something as simple as a filter that was not "on" during the reconstruction; I would assume thay would have caught that right away and told me.  The Kodak support engineers did tell me to use lower kV to improve the image quality; I'm not sure about that!!!! 

LarsBonitz
Offline
Last seen: 35 weeks 4 days ago
Joined: 06/30/2009
poor K9500 images

In my experiences the way to a better quality is not using low radiation, it is using higher radiation. Correcting my CBCT images I using sometimes the new version of Photoshop. Photoshop can now read and manipulate Dicom data. The mass of filters, programmable filters ans Add-ons is huge, so you can find a possibility to enhance the quality of your picture.
It would be a small step to implement some filters to the Kodak software but there are different interest in detail. In my own experiences it was the same.