Susceptibility artefact correction using dynamic graph cuts: Application to neurosurgery

Abstract

Echo Planar Imaging (EPI) is routinely used in diffusion and functional MR imaging due to its rapid acquisition time. However, the long readout period makes it prone to susceptibility artefacts which results in geometric and intensity distortions of the acquired image. The use of these distorted images for neuronavigation hampers the effectiveness of image-guided surgery systems as critical white matter tracts and functionally eloquent brain areas cannot be accurately localised. In this paper, we present a novel method for correction of distortions arising from susceptibility artefacts in EPI images. The proposed method combines fieldmap and image registration based correction techniques in a unified framework. A phase unwrapping algorithm is presented that can efficiently compute the B0 magnetic field inhomogeneity map as well as the uncertainty associated with the estimated solution through the use of dynamic graph cuts. This information is fed to a subsequent image registration step to further refine the results in areas with high uncertainty. This work has been integrated into the surgical workflow at the National Hospital for Neurology and Neurosurgery and its effectiveness in correcting for geometric distortions due to susceptibility artefacts is demonstrated on EPI images acquired with an interventional MRI scanner during neurosurgery.

Simulated field maps for susceptibility artefact correction in interventional MRI

“…The entry points of the simulation are nondistorted preoperative T1w and intraoperative T1w and T2w images A scalar displacement map was calculated using Eq. 1 and was converted into a dense displacement field along the PE direction [7]. Identical displacement was applied to all the EPI images in each DW-MRI data set by resampling the images with cubic spline interpolation using the resampling utility from the NiftyReg package [17].…”

“…The phase-unwrapping was performed using a method detailed in [7], which uses a Markov random field (MRF) formulation to enforce spatial smoothness in the estimated true field map. Since the recovered phase difference necessarily had an arbitrary constant component, we demeaned map based on the voxels inside the brain mask.…”

“…This approach was recently used in a clinical iMRI study and facilitated a reduction in the intraoperative susceptibility artefact [7]. The acquisition process uses the gradient-recalled echo pulse sequence [14], which measures the spin phase evolution between a pair of echoes.…”

“…Techniques for the necessary phase-unwrapping to recover the true continuous field map are presented, e.g. in [12] and [7]. However, an accurate estimation of the true field map becomes difficult near the resection air-tissue boundary.…”

“…Such correction has been shown to reduce tensor fitting errors in DW-MRI data sets [15] and has been used as part of a joint correction scheme in [7]. However, an accurate estimation of the distortion near the resection margin is challenging due to the low spatial resolution of EPI, the heavy compression of voxels and the severe signal dropout that fades and blurs the resection margin and its neighbourhood in the b0-DW-MRI image.…”