Intensity inhomogeneities can obscure areas of interest and are problematic for MR image segmentation algorithms. An efficient approach to estimating these inhomogeneities by computing a calibration factor that is a function of an estimated bias field from a series of calibration scans is disclosed. This enables correction of T.sub.1W and T.sub.2W images by reducing inhomogeneities without the need to map T.sub.1 and T.sub.2. Because of interest in the shape of the intensity inhomogeneities a limited number of flip angles yield satisfactory performance. Additionally, an RF transmit field B.sub.1.sup.+ can be estimated and inhomogeneities resulting from the B.sub.1.sup.+ field reduced from the MR image using the estimated B.sub.1.sup.+ field.