Question on adapting the toolkit for a custom phantom with different specifications #10
Description
Hello,
First, thank you for making this valuable LG-CHO toolkit publicly available. It's an excellent resource for the medical imaging community.
I have a question regarding the toolkit's applicability. I am working with a physical phantom that is conceptually similar to the MITA-LCD phantom, in that it features circular, low-contrast inserts within a uniform background. However, my phantom has different physical specifications:
The overall dimensions of the phantom are different.
The diameters of the low-contrast inserts are different from those in the MITA-LCD phantom.
The materials used result in different Hounsfield Unit (HU) values for both the background and the inserts.
My question is: Is it feasible to adapt the LCD_CT toolkit to analyze images from my custom phantom?
Specifically, I would be grateful for any guidance on the following points:
Are there specific configuration files or hardcoded parameters within the MATLAB scripts that I would need to modify to define the new geometry (e.g., insert sizes, locations) and contrast levels (HU values)?
Given that the signal size and contrast are different, would you recommend re-tuning the Laguerre-Gauss channel parameters for the model observer to ensure optimal performance, or are the default parameters generally robust enough?
Is there any documentation or advice you could offer for users looking to apply this methodology to a non-standard phantom?
Thank you for your time and any help you can provide.