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 2015

SPM June 2015

Options

Subscribe or Unsubscribe

Subscribe or Unsubscribe

Log In

Log In

Get Password

Get Password

Subject:

Re: How to create a mean of two images - and tell SPM about it?

From:

"H. Nebl" <[log in to unmask]>

Reply-To:

H. Nebl

Date:

Tue, 9 Jun 2015 17:46:27 +0100

Content-Type:

text/plain

Parts/Attachments:

Parts/Attachments

text/plain (17 lines)

Dear Andre,

It could be problematic to construct new volumes based on the fXX files. As you encounter motion it is likely subjects don't just jump back to their original position (in which averaging across e.g. vol. 1 and 3 should be sufficient) but that there's a relevant displacement between these two. The constructed new volume might thus be a rather poor replacement as it would be smeared due to motion between the two volumes it's based on. Solutions:
1) As you go with Realign & unwarp anyway you could simply turn to the uf files, the estimated motion parameters have already been applied to the voxel data during the reslice step, so no such problems arise when averaging across different volumes. This is was the Artrepair toolbox does automatically, based on resliced data it allows to generate a new interpolated volume from the neighbouring "good" ones.
2) If you want to work with the (raw) fXX files you could go with separate realignment steps to obtain a mean image based on e.g. vol. 1 and 3, the resulting meanfXX image would then serve as a replacement for vol. 2 and so on.

There are some limitations though. 
3) Often there's fast head motion for several volumes, interpolated data might be no good replacement then.
4) It's questionable whether replacement data can be treated like "correct" data at all. With Artrepair one would usually add a single, binary regressor to the model with one value for the "correct" volumes and another value for the replacement volumes. In case the latter is always the same (e.g. because you don't interpolate corrected data from neighbouring volumes but replace them with a mean image) this should be sufficient. If they are not it would be imperfect, as you try to account for different volumes with a single predictor value, thus just accounting for something like the "average badness". Very likely you should "deweight" these volumes from analysis based on a series of dummy regressors.
5) With a correction as described in 1) and 2) the realignment parameters still reflect motion within the original data, including severe fast motion for bad time points. Thus you try to account for something which is no longer present in exactly this way, or put it differently, you don't properly account for those (head motion) effects which are present in the corrected data. I think there was a methods paper in the context of resting state fMRI / ICA, suggesting that you might (re)introduce artefacts this way.

As a consequence I would just go with the default preprocessing pipeline and add a series of dummy regressors to the design matrix, which corresponds to the motion censoring/"temporal mask" approach described in Siegel et al. (2014, Hum Brain Mapp, https://dx.doi.org/10.1002/hbm.22307 ), as it's a simple but probably quite effective technique. If you prefer to look at the data without Artrepair you can load the rp file with e.g. Excel, SQRT((A2-A1)^2+(B2-B1)^2+(C2-C1)^2+(65^2*((D2-D1)^2+(E2-E1)^2+(F2-F1)^2))) should do it to get their "fast motion" estimate between subsequent time points. A threshold of 0.5 mm/TR might be a reasonable starting point to detect occasional head jumps in otherwise stable subjects, usually you can also detect signal artefacts in the corresponding volumes. To stay on the safe side also deweight the pre- and succeeding volumes (which Artrepair does automatically). Alternatively you could try ART https://www.nitrc.org/projects/artifact_detect/ (fast motion is estimated somewhat differently, thus other thresholds would be required).

Best

Helmut

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