09 Sep 2007

I’ve been working with SPM for some years now, and have decided it’s time for a change. I’m very close to developing some kind of Stockholm syndrome towards SPM, actually defending SPM “features” that obviously just do not make sense.

FSL from the FMRIB group in Oxford is one of the alternatives, also open source, but without the need for an expensive Matlab licence, running nicely on my Mac under X11 and with a course last week in Cardiff (UK).

FSL practicalTBSS

Having been on the SPM course a few years ago, it’s hard not to try to compare the two. Both courses were very well organised, and had an extensive course programme covering fMRI and analysis basics. While the SPM course was very much focused on the analysis software itself (i.e. SPM), the FSL course might just as well have been called general fMRI course, as the FSL tools were presented very much secondary to the general fMRI analysis steps. Although I did think this less self-promoting attitute was very pleasant, I would have liked to hear a bit more about the underlying algorithms implemented in the FSL tools, if only to be able to compare with SPM. The lectures were almost invariably excellent, but maybe a little basic for those with fMRI experience (unlike the SPM course, for which experience with both fMRI and SPM really is required). Much more interesting for the “more experienced” were the twice a day practicals, guiding us through the multitude of quirkily named FSL analysis tools. With many of the easily approachable FSL gurus around, there was ample opportunity to get as much out of this course as we wanted.


In conclusion: very much recommended, both for the beginner and more experienced fMRI researcher.

And, of course, what place better to have a neuroimaging course than where the local beer’s called “Brains”!

Fully PhotoShop-free pictures on Marion’s photos.