A phantom study of new bias field correction method combining N3 and KHM for MRI imaging

Signal inhomogeneity in MRI can influence significantly automatic data processing like segmentation, etc. or even affect the diagnostic procedure. In this work, a new method of intensity nonuniformity correction is presented. Our idea was to replace FCM clustering by k-harmonic means in the method that uses a standard N3 correction procedure. The algorithm was tested with MRI dataset acquired from a phantom object using a breast MRI coil to simulate real conditions during the study. Results were compared with five other methods using two indexes - integral uniformity and standard deviation of the signal inside the object. For the presented and improved method, the lowest integral uniformity and the reasonable low signal deviation were obtained.