(2) Important question with no easy answer. First you must find out if your shift e.g. >1mm at 90/270 is due to patient shift or due to couch run out. I.e. your machine QA should give an answer to the amount of couch runout when rotating from 0° to 90/270. If this value is known, then you can test AlignRT with a head phantom where you know that you have no patient movement. Here the delta should be no larger than the known geometrical couch runout. I did these kind of tests both with a head phantom and 2 volunteer RTTs. Translational deltas where <0.5mm for the full range from 90 to 270 for both phantom and volunteer. rotational deltas where up to 0.4°. So I don’t have any indication for artifacts of AlignRT at extreme couch angles.