Thank you for your reply! We use breast board with couple degree tilt. We have also noticed that in a tensed patient position typically both sternum and the arm are too cranially in the setup images in contrast to vertebra. In some cases we get this information from AlignRT as pitch, some cases not (therefore daily Offline review is still in use). In some cases despite we get this information from AlignRT, it is difficult to correct this pitch. Is it in those cases because it is difficult to copy in the quick patient setup the DICOM position that is acquired at the moment patient is naturally relaxed with the couple minutes delay time? Seems like the situation is inversed the same in the BH cases: if FB VRT/DICOM surface is acquired at the position patient has just taken several BH:s, it may be diffucult to copy this tensed position in the setup process. These kind of things we have focused on a lot, but there is still work left with these things.