In the past clinical scanners had relatively large fringe fields, and when two or more scanners were sited near each other in a facility, some consideration had to be given to the polarity of each magnet. In these situations magnetic fringe fields and interactions were minimized by configuring the magnets so that the unlike poles faced each other. In the modern era with self-shielded scanners, this is not a significant issue for most clinical sites. However, with very high field (7+ Tesla) human and animal scanners now clustered together in research facilities, pole orientation may still be a planning consideration.
Advanced Discussion (show/hide)»
For a single scanner it generally doesn't matter which end is the north pole, or whether Bo goes from front to back or from back to front. The only exception might be when EKG gating is used. As described in a later Q&A, the recorded EKG waveform is altered by the presence of ionic flow in the descending aorta, a phenomenon known as the magnetohydrodynamic (MHD) effect. The direction of Bo will affect the nature of this EKG artifact.
Which way does the main magnetic field point in an MR scanner?