Dear all,
We are running a preprocessing Batch script on fMRI data from 5- and 7-year-old children. The study includes two runs with 3 tasks in each run. After running the script, however, there is consistently an issue whereby the normalised and smoothed functional data do not align well with the anatomical image. Using CheckReg, we notice that the blue crosshairs do not correspond to the same region in the anatomical and functional data (please see the attached images for Subject 1 and Subject 2 in this link: https://drive.google.com/drive/folders/1nzMAoSJ1aWIf_SsN_trbyPYr-x2ORZX6?usp=sharing ). The issue persists whether the script is run for one task alone or all three tasks simultaneously and is present across subjects.
We are using a template from 7-year-old brains and the preprocessing pipeline includes the following steps, in this order: (1) Calculate Fieldmap;
(2) Realignment; (3) Slice timing; (4) Coregistration; (5) Segmentation; (6) Deformation; (7) Normalisation; (8) Smoothing
The script named “NSY_all_preprocessing_2runs” can be accessed here (https://drive.google.com/drive/folders/1nzMAoSJ1aWIf_SsN_trbyPYr-x2ORZX6?usp=sharing).
The issue does not seem to be present in the realigned or slice timing images, only in the normalised and smoothed images. Roughly speaking, for Subject 1 the functional data would need to be rotated slightly forward to align well onto the anatomical data. Might this be related to the artefact at the top of the skull of subject 1? This same kind of artefact also seems to be present in subject 2 but in a more spread-out way.
We would appreciate any suggestions on how to resolve this issue.
|