This chapter introduced the basic principles of magnetic resonance imaging and illustrated the necessity of RF correction for reducing unwanted intensity variation in MR images. Characteristics of MRI data, important to the intracranial boundary detection method developed in this thesis, were presented. The intracranial boundary detection technique can capitalize on the fact that background noise in uncorrected MRI data is characterized by a Rayleigh distribution and that brain voxel intensities, particularly in PD-weighted scans, are approximately normally distributed.