Our lab is doing fieldmap correction for task-based fMRI analyses using SPM12. One participant is missing a field map. A colleague recommended creating an average of all other fieldmaps and using this average for the participant. Another colleague recommended using the fieldmap of another participant in the study for this participant. (The participant had a sibling who did the task immediately after him that same day and he could use his fieldmap.) Does the group have recommendations for how to go about addressing this?
Relatedly, the colleague who recommended creating an average fieldmap recommended betting (skull striping) the magnitude image in order to make the average. However, we previously had not betted the magnitude image, per recommendations from this listserv. Does anyone know if making an average without betting is acceptable or problematic?