Hi Doug,
The the subject data is uploaded, however it seems that we have figured out the issue - so you may not need to look at it. Below is what we think the problem and its solution are. Let me know if it makes sense.

the reconstruction scripts at our campus imaging facility have two flaws: [1] there's is no orientation information in header of either Analyze or NIFTII outputs; [2] the Analyze format is non-standard (RAS instead of LAS).
As a result the first mri_convert step which converts our file to .mgz format assumes that the orientation is neurological resulting in a flip of left to right (albeit with an error message saying that it did not have that info).
this flipping can only be seen when comparing images in tkmedit and SPM, which should have been inveresed, but were not.

We think that in order to avoid this in future we should add a --in-orientation ras flag to mri_convert. Is that correct?

The other question is - now that we have processed some 50 subjects, is there a way of flipping the results so that we wont have to start from scratch (this is a longitudinal study and participants will be scanned again).?

Thanks
Ilana

--

******************************
Ilana Hairston ,PhD
Research Assistant Professor
4250 Plymouth Rd. (rm 2439)
Psychiatry Department, Addiction Research Center
University of Michigan, Ann Arbor MI 48109-2700
Tel: 734-232-0232; Mobile: 650-906-8449

********************
Reminder: Women do 66% of the work worldwide, earn 10% of the income, and own 1% of the world's property