JiscMail Logo
Email discussion lists for the UK Education and Research communities

Help for SPM Archives


SPM Archives

SPM Archives


SPM@JISCMAIL.AC.UK


View:

Message:

[

First

|

Previous

|

Next

|

Last

]

By Topic:

[

First

|

Previous

|

Next

|

Last

]

By Author:

[

First

|

Previous

|

Next

|

Last

]

Font:

Proportional Font

LISTSERV Archives

LISTSERV Archives

SPM Home

SPM Home

SPM  June 2019

SPM June 2019

Options

Subscribe or Unsubscribe

Subscribe or Unsubscribe

Log In

Log In

Get Password

Get Password

Subject:

Re: small volume correction procedure

From:

MRI More <[log in to unmask]>

Reply-To:

MRI More <[log in to unmask]>

Date:

Thu, 27 Jun 2019 13:53:44 +0100

Content-Type:

text/plain

Parts/Attachments:

Parts/Attachments

text/plain (17 lines)

Dear Ariadna,

Starting with some general remarks: The concept of a small volume correction is to restrict the analysis to regions with a-priori hypotheses instead of running a whole-brain analysis, the motivation is to end up with a more liberal threshold due to smaller volume. As one would have to decide about thresholding in advance, one would have to incorporate *all* a-priori regions. This is very important. In practice, I have never seen a paper proceeding like that. Rather, it's whole-brain analyses, showing sig. activations in certain a-priori regions, followed by SVC for a-priori regions that failed to reach sig. in the whole-brain analyses. This is bad science. 1) In this context, the decision to conduct a SVC means to lower the threshold only after one has run some analyses (if it had been sig. on whole-brain, there would obviously have been no need for SVC), which is a flaw. 2) The threshold is lowered for *some* regions only, which is also a flaw. One could argue that this is a minor issue, if a region shows up whole-brain, it should also be sig. in a SVC. However, the extent of the activations regarded as sig. might be (much) larger, which could have certain implications, plus, when considering all a-priori regions for SVC, the resulting threshold might be too high so that the initial sig. SVC results for select a-priori regions are no longer sig. It's worth a note that the SVC threshold does not reflect the volume but the RESEL count, which is affected not only by volume but also by surface and smoothness of the residuals. Several small spheres could already result in a relatively large surface.

Concerning 1.: In principle, one can think of two options*. Combine all a-priori regions into a single volume (mask) or run several SVC restricted to single regions, accounting for the number of tests (e.g. in case of five a-priori regions and a chosen initial voxel threshold of .001, go with .001/5). The results will not be identical as the a-priori regions likely differ in volume, surface, smoothness. As an example, think of some low-level vision experiment in which we are only interested in occipital lobe and LGN. The occipital lobe is much larger than the LGN. In case of a SVC on cluster level covering both regions, a cluster would probably have to be quite large to reach sig. in LGN. It might be wiser to run two separate SVCs with an initial voxel threshold .001/2. Whether to go this or that way, well, there are always drawbacks.
* Actually, there's a third one. Combine all a-priori regions into a single volume (mask) and specify this as an explicit mask when setting up your GLM, then run the model and plot and report the results as you would usually do. The whole-brain analysis is now an "a-priori region analysis". These results will not be identical to SVC results derived from a single volume (mask), as some parameters are approximated from the whole-brain volume, that is also from voxels that have never entered the "a-priori region analysis".

Concerning 2: This is correct for SVC on cluster level. Note, however, that SVC is often conducted on voxel level, probably more often than on cluster level - usually resulting in inconsistent threholding = starting with whole-brain analyses with cluster correction, followed by SVC with correction on voxel level. For SVC on voxel level, the procedure is a bit intricate https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1504&L=SPM&D=0&P=450761 .

In general, I would however disadvise from running SVCs. The idea of a SVC is reasonable, but the way it is used turns out to be problematic to such an extent that I would just forget about it. I would rather go with region-of-interest analyses e.g. with Marsbar, plot the average estimates for the different regions and conditions and conduct the analyses based on the extracted parameters. This should be way more informative, after all you're looking at a-priori regions. Please note that this type of ROI analysis does not answer the questions that a SVC could answer, as SVC is about sig. voxels / clusters within a pre-defined volume and not about the average.

Best regards

Helmut

PS: I'm the author of the old message that you had referred to, in the meantime I use a differen mail address for the forum, so don't be surprised.

Top of Message | Previous Page | Permalink

JiscMail Tools


RSS Feeds and Sharing


Advanced Options


Archives

October 2019
September 2019
August 2019
July 2019
June 2019
May 2019
April 2019
March 2019
February 2019
January 2019
December 2018
November 2018
October 2018
September 2018
August 2018
July 2018
June 2018
May 2018
April 2018
March 2018
February 2018
January 2018
December 2017
November 2017
October 2017
September 2017
August 2017
July 2017
June 2017
May 2017
April 2017
March 2017
February 2017
January 2017
December 2016
November 2016
October 2016
September 2016
August 2016
July 2016
June 2016
May 2016
April 2016
March 2016
February 2016
January 2016
December 2015
November 2015
October 2015
September 2015
August 2015
July 2015
June 2015
May 2015
April 2015
March 2015
February 2015
January 2015
December 2014
November 2014
October 2014
September 2014
August 2014
July 2014
June 2014
May 2014
April 2014
March 2014
February 2014
January 2014
December 2013
November 2013
October 2013
September 2013
August 2013
July 2013
June 2013
May 2013
April 2013
March 2013
February 2013
January 2013
December 2012
November 2012
October 2012
September 2012
August 2012
July 2012
June 2012
May 2012
April 2012
March 2012
February 2012
January 2012
December 2011
November 2011
October 2011
September 2011
August 2011
July 2011
June 2011
May 2011
April 2011
March 2011
February 2011
January 2011
December 2010
November 2010
October 2010
September 2010
August 2010
July 2010
June 2010
May 2010
April 2010
March 2010
February 2010
January 2010
December 2009
November 2009
October 2009
September 2009
August 2009
July 2009
June 2009
May 2009
April 2009
March 2009
February 2009
January 2009
December 2008
November 2008
October 2008
September 2008
August 2008
July 2008
June 2008
May 2008
April 2008
March 2008
February 2008
January 2008
December 2007
November 2007
October 2007
September 2007
August 2007
July 2007
June 2007
May 2007
April 2007
March 2007
February 2007
January 2007
2006
2005
2004
2003
2002
2001
2000
1999
1998


JiscMail is a Jisc service.

View our service policies at https://www.jiscmail.ac.uk/policyandsecurity/ and Jisc's privacy policy at https://www.jisc.ac.uk/website/privacy-notice

Secured by F-Secure Anti-Virus CataList Email List Search Powered by the LISTSERV Email List Manager