Correction of magnetic resonance imager intensity inhomogeneities using tissue properties

摘要

A statistical method for correcting for intensity inhomogeneities in the images produced by magnetic resonance imaging equipment. The method uses knowledge of tissue properties and gain inhomogeneities to correct the gain artifact, thereby improving intensity-based segmentation of the volume spanned by the imagery into the various tissue types that are present. The method utilizes the Expectation-Maximization algorithm to simultaneously estimate tissue class and the radiofrequency gain field. The algorithm iterates two components to convergence: tissue classification and gain field estimation. The method combines them in an iterative scheme that yields a powerful technique for estimating tissue class and radiofrequency gain.