Hi James, how are you?
I´m trying to analyze the monthly qa start shot using CR cassettes from Karestrem.
I did collimator (colimador), couch(mesa) and gantry using 2 monitor units and a MLC aperture of 5mm
I don´t understand yet how to adjust the parameters to get the right results.
The collimator and couch are fine, the problem is the gantry.
I realize that the code is using the values between the the shots instead of using the shots.
Do you think I need to change the gantry values?
Thanks
Hi Pedro,
I’m not exactly sure what you mean here. Attaching a picture of your results or your raw image would help. Occasionally, pylinac will invert the image incorrectly (noisy images and artifacts contribute to this). The first thing to do is use a low radius value in the analyze() method. Maybe 0.2.
If the image is finding the “valleys” instead of the peaks, then you’ll have to force invert the image. I should have added a simple “invert” parameter to analyze(), but I’m an idiot and didn’t do it. In your source code for pylinac (in your local pylinac package) you’ll have to change this line (https://github.com/jrkerns/pylinac/blob/master/pylinac/starshot.py#L205) to something like “…=[96, 50, 4]” which will invert the automatic histogram calculation and invert your image.
Hi Pedro,
Did you resolve the problem with gantry starshot ? I am getting runtime error so have you changed the source code ?
Thanks,
Ahmad Amoush
No. I have not changed the code, for me it is too big and complex.
I need to send the dicom files to James.
My best regards