Magnetic Resonance Imaging with non-Cartesian acquisition schemes suffer from blurring artifacts induced by offresonance. Conventional algorithms for off-resonance correction are approximate in nature and assume that the offresonance map is varying smoothly across the object. This assumption is violated in the case of susceptibility related off-resonance which is sharply varying near certain tissue interfaces. Iterative methods have been proposed for sharply varying off-resonance correction but are either computationally intense in the exact form or need approximations like time segmentation to reduce computational cost. Here a new method for accurate and computationally efficient off resonance correction using spatial domain deconvolution is presented.
Mahender K. Makhijani, Krishna S. Nayak