Looking at the analysis of picket fence testing I’m somewhat perplexed. It seems like in DoseLab, Pylinac, and maybe some other packages the analysis focuses on the distance between pickets and the leaf positions relative to the pickets. What appears to be missing is the location of the pickets relative to isocenter.
Is there a reason why ‘absolute’ position of the pickets isn’t a critical parameter?
In pylinac can I use the pickets.dist2cax attribute to assess absolute position reliably? (ie, does it know about imager offsets or does it just use the center of the image, etc)
I think the reason is historical. Absolute position is obviously quite important. The center of the image is a good value to compare to if you trust your imaging panel. With Isocal, it’s dead on, so I had no problem trusting it. If you don’t trust your image center then you could do a 90/270 collimator PF with offset pickets (e.g. MLCs with an asymmetric offset of 2cm and step of 3cm). Then the distance between the pickets can be compared to nominal. You can also test your imager by comparing WL images of the BB vs EPID center, however I did not get results as accurate as I would have liked.