摘要

Shimming is very important for nuclear magnetic resonance (NMR) magnets because image resolution is highly dependent on the homogeneity of the magnetic field. There are two types of shimming: active and passive. Active shimming is done using coils with adjustable current. Passive shimming involves pieces of steel with good magnetic qualities. The steel pieces are placed near a superconducting magnet. They are magnetized and produce their own magnetic field. Additional magnetic fields (produced by coils or steel) add to the original magnetic field of the superconducting magnet in such a way that the total field becomes more homogeneous.
In this paper, we developed a passive shimming method adopting consecutive optimization techniques, i.e., linear programming (LP) and evolution strategy (ES). The LP is relatively fast and mostly guarantees a global minimum for a linear problem, whereas the ES is easy to formulate and can digitize design variables. So we suggested an optimization method combining both the LP and the ES consecutively for passive shimming of NMR magnets.

  • 出版日期2011-1