Hi Lara,
the longitudinal pipeline is designed for adults with the assumption that head size is fixed. We are working on a new pipeline that will allow processing of images from children, but that will take a while. Especially considering that there will be large differences in the brain as well as the head size between age 2 and age 10, I recommend to only use the cross sectional pipeline for this data.
Best, Martin
On 01/13/2016 12:05 PM, Lara Foland-Ross wrote:
Dear Freesurfer experts,
I am working on a longitudinal analysis of cortical structure and am experiencing some errors with registration. I have 3 timepoints for each participant:
time 1: age 2 time 2: age 5 time 3: age 10
Time 1 and 2 are from a 1.5T scanner. Time 3 is from a 3T scanner.
Despite the age range and scanner difference, the registrations in the longitudinal pipeline (I'm using version 5.3) are performing relatively well. However, in comparing the LONG outputs slice by slice, it appears that in a majority of my subjects, there are some minor mismatches in brain anatomy. An example is attached here (minor_reg_error.jpg).
In other instances, there are more substantial registration errors. Again, example attached here (major_reg_error.jpg).
My question is how much mismatch is acceptable? And how can I go about fixing poor registrations in subjects who have significant errors in alignment?
Many thanks in advance, Lara
Lara Foland-Ross, Ph.D. Research Associate Center for Interdisciplinary Brain Sciences Research Stanford University School of Medicine 401 Quarry Road, Room 1356 Stanford, CA 94305-5795
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer