Cascaded slice to volume registration for moving fetal FMRI

Abstract

Motion correction of MRI sequences is a very active area of research. Several postprocessing techniques for volume correction and more recently slice correction have been proposed. Slice motion correction of fMRI data typically involves iterative registration of slices to a target volume. The target volume is usually reconstructed at each iteration using current slice motion estimates, with all possible views of the subject. However, in the presence of large movements, the quality of the reconstruction can be greatly degraded without a good initialization, and this can in turn greatly affect the accuracy and precision of the slice registration. This paper introduces a cascaded motion correction pipeline that uses subsets of data to generate coherent volumes for initial target volume based slice registration in fMRI. The pipeline incorporates distortion correction, frame motion estimation and two levels of slice to volume registration. The proposed method has been tested on moving adult and fetal fMRI data.

3D in utero quantification of T2* relaxation times in human fetal brain tissues for age optimized structural and functional MRI

“…The brain masks used by SLIMMER were based on manually delineated fetal brain segmentations averaged for subjects within age ranges of 2 GW in order to represent the appropriate brain size and shape. Finally, to correct the effects of fetal head motion that occurred between the acquisitions of subsequent slices (between-slice motion), cascaded slice to volume registration was applied (17). The final 3D T2*-w reconstructions were created in 1.0 mm 3 isotropic resolution by applying linear interpolation at the initialization step followed by the iterative optimization process.…”

Detecting default mode networks in utero by integrated 4D fMRI reconstruction and analysis

“…This is because the acquisition is fast enough that within-slice motion artifacts are minimal for most slices even in the case of significant motion. If we employ a slice alignment algorithm to estimate the rigid motion mapping of each slice into a common anatomical coordinate frame (Seshamani et al 2013) and the subject is undergoing large motion, they may in general be distributed in an uneven pattern over this common anatomical space. As signal change over time is critical in fMRI analysis, and each slice is acquired at a distinct temporal point, both the time and location of the slices must be considered.…”

“…They propose methods for slice motion estimation and scattered data reconstruction within a 3D volume. These methods have been previously applied to fMRI data (Bhagalia and Kim, 2008; Seshamani et al 2013) with separate reconstruction of each 3D volume within the time sequence. However, it is well known that fMRI data is serially correlated (Chaari et al 2011) and temporal smoothing is often applied as a post-processing step.…”

Graph-Based Slice-to-Volume Deformable Registration

“…The popular map-slice-to-volume (MSV) method that introduced this idea in the context of functional MRI (fMRI) was presented by [20]. More recently, applications of slice-to-volume registration to the same problem in different contexts like cardiac magnetic resonance (CMR) [5], fetal images [41] and diffusion tensor imaging (DTI) [18] have shown promising results.…”