This notice is part of our continuing AFDI.*
Recently, the AFNI team published our recommendations for quality control in FMRI.
https://www.frontiersin.org/articles/10.3389/fnins.2022.1073800/full
In this paper, we examined several FMRI datasets from open repositories. While we now pretty much expect a left-right flipping error in new data, we were surprised to find some subject brains appeared upside down. The eagle-eyed among AFNI users will see this in two subjects below.
https://afni.nimh.nih.gov/pub/sscc/staff/glend/group5_montage_with_upside_down_subs.sag.png
Since that publication, it has come to our attention from our colleagues “down under” in Australia and New Zealand that this is, in fact, their standard. While they prefer to remain anonymous, they sent us this dataset. Surprisingly, the “upside-downness” is correct for this set.
https://afni.nimh.nih.gov/pub/sscc/staff/glend/aus_group4_montage_mostly_upside_down_subs.sag.png
They are not sure of why this happens, but they describe it as something akin to the coriolis effect seen in water drains, and call it the “boomerang effect”.
QC of the Australian datasets revealed an outlier… Our local Kiwi, Peter Lauren, commented, “this doesn’t surprise me one bit.”