Hi,
I preprocessed resting-state fMRI data using fMRIPrep 25.2.4 with --use-syn-sdc (fieldmap-less SDC).
The B0 fieldmap report shows strong horizontal/vertical striping in coronal/sagittal views (attached).
Is this artifact typical for SyN-based fieldmap-less estimation, or does it indicate a registration issue? Phase-encoding direction was left–right; no actual fieldmaps were acquired.
I have not run fMRIPrep without --use-syn-sdc, in my dataset did not acquire fieldmaps, and disabling SDC would leave severe susceptibility distortions uncorrected (especially with LR phase encoding). However, I can test this if helpful.
The raw DICOMs and reconstructed BOLD timeseries appear normal: no striping is visible in the 4D NIfTI volumes or in standard viewers. The artifacts are only present in the estimated B0 fieldmap (i.e., the desc-anat_fieldmap.svg figure), not in the preprocessed BOLD data itself.
Specifically, the striping appears in the deformation field visualization (coronal/sagittal views of the fieldmap in Hz), which is a known behavior of SyN-based fieldmap-less estimation when using low through-plane resolution (4 mm slices) and constrained registration along the phase-encoding direction.
