Reduction of susceptibility artifacts in functional MRI (fMRI)


Differences in magnetic susceptibility at air-tissue boundaries cause inhomogeneities in the static magnetic field (B0 field). These field inhomogeneities result in geometric distortion and BOLD sensitivity loss in gradient-echo echo-planar imaging (EPI) [1], particular in basal brain areas such as the orbitofrontal cortex (OFC).


Compensation of susceptibility-induced BOLD sensitivity loss

Susceptibility-induced B0 field gradients reduce the BOLD sensitivity in fMRI. The effects can be categorized according to the relative direction of the field gradient with respect to the EPI encoding directions: 1) through-plane/slice select [1], 2) phase encoding [1] and 3) readout direction [2]. Each type will result in a characteristic BOLD sensitivity loss and requires different compensation approaches. We use primarily carefully optimized slice angulation, phase-encoding direction, z-shim and high-resolution acquisition to overcome the BOLD sensitivity loss problem [1], [2] (Figs. 1 and 2). In addition, we have developed a novel approach to shimming, which optimizes the BOLD sensitivity and thus directly the effect of interest in fMRI [3].


Current projects involve high resolution fMRI at 1.5 mm isotropic resolution and variable slice-dependent z-shim.




Fig 1: Improvement in BOLD sensitivity due to increased resolution (1.5mm versus 4 mm) in the readout direction (left-right) in case of dropouts due to susceptibility-induced field gradients in the same direction (modified from [2]. Note the improvement in the U-shaped dropout in the orbitofrontal cortex (OFC).




Fig 2: Improvement in BOLD sensitivity due to optimized slice angulation, PE direction and z-shim (modified from [1]), recovering loss due to susceptibility-induced gradients in the PE and slice select direction.


Correction of susceptibility-induced image distortion

Due to the rather small bandwidth in the phase-encoding (PE) direction EPI suffers from geometric distortion in the PE direction in the presence of B0 field inhomogeneities [4]. We have developed methods for correcting these static geometric distortion based on B0 fieldmaps [4]. In addition, novel approaches allow for compensating dynamic changes in B0 caused by respiration or head movement. These correction methods are widely used and implemented as toolboxes for SPM (e.g., realign and unwarp in SPM).


Current projects focus on alternative methods for dynamic B0 correction based on the EPI phase information and linear modeling based on motion parameters or physiological parameters (such as respiration) [5].

Primary contact

Nikolaus Weiskopf (n.weiskopf «at»



[1]        N. Weiskopf, C. Hutton, O. Josephs, and R. Deichmann, “Optimal EPI parameters for reduction of susceptibility-induced BOLD sensitivity losses: a whole-brain analysis at 3 T and 1.5 T,” Neuroimage, vol. 33, pp. 493-504, 2006.

[2]        N. Weiskopf, C. Hutton, O. Josephs, R. Turner, and R. Deichmann, “Optimized EPI for fMRI studies of the orbitofrontal cortex: compensation of susceptibility-induced gradients in the readout direction,” MAGMA., vol. 20, no. 1, pp. 39-49, Feb. 2007.

[3]        E. Balteau, C. Hutton, and N. Weiskopf, “Improved shimming for fMRI specifically optimizing the local BOLD sensitivity,” Neuroimage, 2009.

[4]        C. Hutton, A. Bork, O. Josephs, R. Deichmann, J. Ashburner, and R. Turner, “Image distortion correction in fMRI: A quantitative evaluation,” Neuroimage., vol. 16, no. 1, pp. 217-240, May 2002.

[5]        C. Hutton, J. Andersson, R. Deichmann, and N. Weiskopf, “Phase informed model for motion and susceptibility (PIMMS),” Hum.Brain Mapp., vol. submitted, submitted.

back to physics pages