Just to follow up, Scott. It looks like MarsBar would work well for
manipulating the functional data once I have defined and traced ROIs,
but now I'm faced with a decision on whether it is the average activity
differences between conditions within the ROI that is of primary
importance (MarsBar), or whether I'm expecting to see spatially
differential activity within an ROI between conditions (SVC). For
relatively small and circumscribed ROIs, it appears that MarsBar would
work best, but for larger less functionally homogeneous ROIs (e.g.,
total hippocampus ROI) from what I've learned SVC may be the way to go.
If anyone with direct experience in either technique can contribute to
the discussion, it would be most appreciated.
Warm regards,
Jeff
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]]
On Behalf Of Jeff Browndyke
Sent: Sunday, March 05, 2006 2:48 PM
To: [log in to unmask]
Subject: Re: [SPM] Non-altas based analyses / ROI mask creation & use
with SPM?
Hey, Scott.
I'll have to re-review the MarsBar site, but I was under the impression
that it works off the assumption of normal (or near normal) anatomy.
Hopefully I'm mistaken about this because it would be nice to have a
well-supported tool on hand to put our methods into place.
I suspect (as with just about anything dealing with imaging analysis)
that we will need to beg, borrow, and steal from many programs and
methods to tweak it just right.
Thanks,
Jeff
-----Original Message-----
From: Scott Hayes
Sent: Sunday, March 05, 2006 2:25 PM
To: Jeff Browndyke
Subject: Re: [SPM] Non-altas based analyses / ROI mask creation & use
with SPM?
Hi Jeff
There's folks w/ more experience than me, but MarsBar (by Mathew Brett
http://marsbar.sourceforge.net/) is a pretty popular ROI toolbox that is
often used w/ SPM2. They've got some nice tutorials, and there's a
listserv for the software as well. I know it will do some of the things
you're looking for (draw an ROI, cross your ROI w/ functional activation
maps, and extract mean activation w/in an anatomically defined ROI), but
I'm not sure if it will do all of them (e.g., "count" # of active voxels
w/in an ROI).
Best,
Scott
Jeff Browndyke wrote:
>Listmates,
>
>I'm in the process of deciding how to best implement a non-atlas based
>approach to analyzing our fMRI data, and I would like to learn from the
>experience of others as to which materials and methods would be best
for
>our needs. Since the patient population for which we have fMRI data is
>fraught with the presence of brain atrophy (some quite significant), it
>doesn't seem appropriate to warp to a common SPM template. Rather, I
>would like to draw anatomically defined masks for particular regions of
>interest and examine the activity characteristics within these regions
>(e.g., ratio of activated voxels to total voxels, mean activation of
>voxels in ROI, etc.). The ROIs would also provide us the valuable data
>of accounting for volumetric differences in these regions among the
>subjects, which can then be used to correct for activity/atrophy
>interactions.
>
>What tracing software for the creation of anatomical masks best works
>with SPM2 or SPM5? And, does anyone know of a primer, good methods
>paper, or workshop for how to implement this anatomical ROI/non-atlas
>approach in SPM2 or SPM5?
>
>My previous SPM experience has been confined to normal anatomy, so what
>we're hoping to do is a bit foreign right now. I want to do the right
>thing with this data and warping to the SPM template doesn't seem to be
>it. I know manual tracing will be a major pain, but I don't see any
>other way to work around the atrophy/function issue.
>
>Regards,
>Jeff Browndyke
>
>
|