Accelerated point spread function mapping using signal modeling for accurate echo-planar imaging geometric distortion correction.

Echo-planar imaging is a fast and commonly used magnetic resonance imaging technique with applications in diffusion weighted and functional MRI. Fast data acquisition in echo-planar imaging is accomplished by the extended readout, which also introduces sensitivity to off-resonance effects such as amplitude of static (polarizing) field inhomogeneities and eddy-currents. These off-resonance effects produce geometric distortions in the corresponding echo-planar images. To correct for these distortions, an acceleration of point spread function (PSF) acquisition using a special sampling pattern is presented in this work. The proposed technique allows for reliable and fully automated distortion correction of echo-planar images at a field strength of 3 T. Additionally, a new approach to visualize and determine the distortions in a hybrid (x, y, kPSF) three-dimensional space is proposed. The accuracy and robustness of the proposed technique is demonstrated in phantom and in vivo experiments. The accuracy of the presented method here is compared to previous techniques for echo-planar imaging distortion correction such as PLACE.