Dear SPMers,
we have the following problem: we have 30 functional and one structural volumes
of the same subject; the structural volume covers the whole brain while the
functional slices are fewer and cover a smaller region.
We perform the statistical analysis of the realigned and smoothed functional
images and obtain some activations. Now we wonder how to map the activation over
the structural slices. We think we can follow three routes:
1) Since we are not comparing different subjects, we can simply use the overlay
button and choose the structural slices.
2) We use the coregister option with a functional image as target and the
structural volume as object; make the analysis of the original functional
volumes and use the overlay button which will load the coregistered anatomical
volume.
3) We use the coregister button with the anatomical volume as target, and the
functional ones as object and make the statistical analysis of the coregistered
functional images, overlaying the activation on the untouched anatomical slices.
Now we did these three things and the result is that point 2) and point 3) give
similar (but not identical) results while point 1) shows the activations on very
different (significantly higher) anatomical slices.
We have many troubles about all this: is it correct to coregister volumes
covering different brain regions? If so, what are the reasons of so different
results? Is it true that in this situation the safest strategy is point 1) ?
Thanks in advance,
Michele Piana
INFM -Genova
Italy
[log in to unmask]
|