Template-based field map prediction for rapid whole brain B₀ shimming

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging (MRI)
• Biophysics

Background:

• Acquiring field maps for MRI shimming is time-consuming.
• Optimizing shim settings is crucial for high-quality MRI.
• Existing methods often require lengthy acquisition times.

Purpose of the Study:

• To develop and evaluate a fast template-based method for predicting magnetic field maps.
• To achieve near-optimal shim settings without direct field measurement.
• To reduce the time spent on acquiring field maps in MRI protocols.

Main Methods:

• Utilized a template-based prediction approach using a large database of human brain field maps.
• Incorporated subject-specific structural information from quick localizer scans.
• Compared the shimming performance of the template-based prediction against other fast shimming methods.

Main Results:

• Static B0 shimming using predicted field maps performed comparably to using individually measured field maps.
• The proposed method achieved a residual field standard deviation of 59 Hz in the brain at 7T.
• This is a significant improvement compared to 176 Hz with no subject-specific shim.

Conclusions:

• Shimming based on predicted field maps is a feasible and effective technique.
• The template-based prediction method offers a time-efficient alternative for MRI shimming.
• Further improvements in prediction accuracy are possible through template refinement.