Print

Print
You'd need to test that you get the same statistics, which is easy enough to do with one map.

Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Research Fellow, Department of Neurology, Massachusetts General Hospital and
Harvard Medical School
Postdoctoral Research Fellow, GRECC, Bedford VA
Website: http://www.martinos.org/~mclaren
Office: (773) 406-2464
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED
HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is
intended only for the use of the individual or entity named above. If the
reader of the e-mail is not the intended recipient or the employee or agent
responsible for delivering it to the intended recipient, you are hereby
notified that you are in possession of confidential and privileged
information. Any unauthorized use, disclosure, copying or the taking of any
action in reliance on the contents of this information is strictly
prohibited and may be unlawful. If you have received this e-mail
unintentionally, please immediately notify the sender via telephone at (773)
406-2464 or email.

On Tue, Mar 10, 2015 at 8:53 PM, Andy Yeung <[log in to unmask]> wrote:
Dear all,

Thank you Matthew for directing me to Marsbar. The conceptual differences between ROI and SVC are now clear to me, the former assumes homogeneous activation while the latter assumes a heterogeneous one. Thank you. And thanks Donald for showing me the different steps used in ROI and SVC.

May I paste an answer from Researchgate.com here. Could anyone confirm the answer is accurate?
That practically using ROI inbuilt in SPM or through WFU Pickatlas produce the same results as SVC in SPM.

Best,
Andy

Forwarded message begins:

Hi Andy,

I think it should technically make no difference.

Do you have the WFU Pick Atlas installed? As far as I remember it is this toolbox that adds the ROI dialog into the SPM GUI, and restricts the search space beforehand, while the built-in SVC does it afterwards.

Both should compute a FWE-corrected p value based on the resels in this restricted search volume. Try out both with the same ROI definitions, and look at the Voxels, FWHM etc. information at the bottom of the table - they should look the same.

Concerning the (mis)use of ROI in general, the following article by Russ Poldrack on the use of ROIs is also quite helpful.

http://scan.oxfordjournals.org/content/2/1/67.abstract

(See also the more recent discussion in the Poldrack/ Mumford/ Nichols book: http://www.fmri-data-analysis.org/)

Hope this helps,

Marcel



On Sun, Mar 8, 2015 at 3:45 AM, Matthew Brett <[log in to unmask]> wrote:
Hi,

On Sat, Mar 7, 2015 at 3:41 AM, MCLAREN, Donald
<[log in to unmask]> wrote:
> There are quite different:
> (1) ROI Analysis - this typically means that you select your ROI, extract
> the data, and then run the statistics on the ROI mean/eigenvalues.
> (2) SVC Analysis - this is a voxel-wise analysis within a restricted area of
> the brain. The maps don't change, but the corrected p-values will change
> because you have restricted the search region.
>
> With an ROI analysis, you can test multiple ROIs. With SVC, you have to test
> all regions simultaneously or correct for the number of times you perform an
> SVC.
>
> The math underlying SVC is the same as the whole brain analysis, except the
> search volume is reduced.

There's a marsbar faq topic that might be helpful:

http://marsbar.sourceforge.net/faq.html#should-i-use-marsbar-roi-analysis-or-small-volume-correction-svc-in-spm

Cheers,

Matthew