External Email - Use Caution
Hello FS experts,
I am using longitudinal processing pipeline ( https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalProcessing) to calculate the cortical volume (CV) over a course of treatment (two conditions: pre condition and post condition). In fact, I am interested in comparing normalized CV (NCV) i.e. raw CV/ ICV between pre- and post-condition.
(1). Now, longitudinal pipeline assumes that there is no change in ICV so it gives me identical ICV for pre and post condition. That way, when I compared NCV between pre and post condition, it gives me differences at significant level of 0.16. (2). However, when I checked ICV (from cross-sectional pipeline i.e. each time-point separately) and compared between pre and post-conditions, ICV values are different. Its possible that these values could be different because of treatment (may be !). When I compared NCV between pre and post condition (calculated by dividing raw CV from longitudinal pipeline with ICV values form cross-sectional pipeline), it gives me significant differences at 0.02.
Could you please share your thoughts on this i.e. (a) whether I can use approach 2 or (b) if I am using longitudinal pipeline, dividing raw CV from longitudinal pipeline and dividing by ICV calculated from cross-sectional is in correct way to do this analysis?
Thanks.