SBRT 2016

Observations

 The distance to the center of the magnet seems to be an important factor for geometric distortion in the CC direction. It is even more important than whether a T1w or T2w sequence is used  The 3D geometrical correction seems to only work on the T1w scan. For this sequence the CC-error is reduced to a level below the slice spacing (4 mm)  For the T2w scan the 3D algorithm does not seem to work: the CC-error can still be as large as 7 mm for points far away from the magnet center