headers
Murphy, Eric | 9 Oct 23:15

timecourse functions

Hi Matthew,

 

I have some questions about the two timecourse functions on Marsbar, and I was hoping you could give me a straightforward explanation on a few things.

 

I am testing out Marsbar on some very simple checkerboard data, 20 seconds rest, 20 seconds checker, repeated once.

 

When I try to create a fitted event timecourse or a FIR timecourse, is it only possible to create the timecourse for a single event, or can I see both sets of “on” and “off” on one continuous graph?  If it is only possible to see one event at a time, and I have two runs of the experiment, would it make more sense to model the two runs separately and then average them, or to combine the two checker conditions in one “event”?

 

Finally (for now) what exactly is the difference between the fitted event timecourse and the FIR timecourse options?  They both ask for the same information, and the graphs look identical.  Am I missing something here? 

 

Thanks for the help.

 

Eric Murphy, M.A.

Research Assistant

Dept. of Psychiatry & Behavioral Neuroscience

Wayne StateUniversityMedicalSchool

emurph <at> med.wayne.edu

(313) 577-1838

 

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Marsbar-users mailing list
Marsbar-users <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/marsbar-users
Permalink | Reply | 1 comment |
headers
Matthew Brett | 10 Oct 12:51
Picon
Gravatar

Re: timecourse functions

Hi,

> When I try to create a fitted event timecourse or a FIR timecourse, is it
> only possible to create the timecourse for a single event, or can I see both
> sets of "on" and "off" on one continuous graph?  If it is only possible to
> see one event at a time, and I have two runs of the experiment, would it
> make more sense to model the two runs separately and then average them, or
> to combine the two checker conditions in one "event"?

It's the usual SPM thing, you model an event type - usually - so one
regressor for the checkerboard on, with two events in it.  You can
only plot and event type, and not the events within it.  You can of
course model both events with a separate set of regressors, and then
you can plot the individual events.

> Finally (for now) what exactly is the difference between the fitted event
> timecourse and the FIR timecourse options?  They both ask for the same
> information, and the graphs look identical.  Am I missing something here?

It depends on your original SPM model.  If you used an FIR
basis-function model for SPM then the FIR and fitted time courses will
be very similar.  If you used - say - a canonical HRF for your SPM
basis functions, then the fitted time course will reflect this more
limited model compared to the FIR model.

Best,

Matthew

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Permalink | Reply | 0 comments |
headers
Murphy, Eric | 12 Oct 20:44

saving previous ROI data

Hi Matthew,

 

I am experiencing a bit of confusion with the structure for saving information on Marsbar, and I was hoping you could clear something up for me.

When I run an analysis, say a raw data timecourse, I end up with a nice graph of the timecourse.  Since I am more interested in the specific values on the graph, I export the data to an Excel file, where I can see the numbers.  I have the specific data I need saved, and everyone is happy.  However, when I go to do my next analysis, Marsbar asks me if I want to save my previous ROI data, and I have no idea what to do.  What exactly am I being asked to save at this point?  The graph?  And if I do save it, or some other analysis, how do I bring it back up again to look at without going through the entire analysis again? 

I know these are all very basic questions, but I am not finding the saving methods very intuitive, and the tutorial doesn’t really go into it as well as I would hope.

Thanks for your continued and future help.

 

Eric Murphy, M.A.

Research Assistant

Dept. of Psychiatry & Behavioral Neuroscience

Wayne StateUniversityMedicalSchool

emurph <at> med.wayne.edu

(313) 577-1838

 

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Marsbar-users mailing list
Marsbar-users <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/marsbar-users
Permalink | Reply | 1 comment |
headers
Matthew Brett | 13 Oct 12:41
Picon
Gravatar

Re: saving previous ROI data

Hi,

> When I run an analysis, say a raw data timecourse, I end up with a nice
> graph of the timecourse.  Since I am more interested in the specific values
> on the graph, I export the data to an Excel file, where I can see the
> numbers.  I have the specific data I need saved, and everyone is happy.
> However, when I go to do my next analysis, Marsbar asks me if I want to save
> my previous ROI data, and I have no idea what to do.  What exactly am I
> being asked to save at this point?  The graph?  And if I do save it, or some
> other analysis, how do I bring it back up again to look at without going
> through the entire analysis again?

You are being asked to save the data as the marsbar data .mat format -
so it is nothing to do with the plot that you saved.  You might do
this if extracting the data took a long time, and you wanted to save
yourself that time by just loading the extracted data directly...

Best,

Matthew

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
Permalink | Reply | 0 comments |
headers
Matthew Brett | 13 Oct 20:49
Picon
Gravatar

Minor new devel release

Hi,

Just to let y'all know I've put up a minor new release,
marsbar-devel-0.41, to fix a few bugs people have spotted.  I've also
updated the download link on the home page to make it more obvious
where to get the devel release and other marsbar packages.  As usual,
please let me know of any problems,

Best,

Matthew

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
Permalink | Reply | 0 comments |
headers
Scott, David | 16 Oct 20:36
Picon
Favicon

Different raw data with one vs. multiple subjects

Hello marsbar users,

I have come across a very strange issue using marsbar to extract raw data from image series. If I extract raw data (as described in the FAQ) from an ROI when applied to images from several different subjects, I get one set of values for each image. However if I apply this ROI in the same manner to these subjects one at a time, I get a very different set of values from these images. I am not using any scaling or anything like that so I don't see why these "raw" values would be changing depending on how many of the images I extract from. Has anyone come across this, or could offer some advice?


Thanks in advance!

Dave Scott

----------------------------------------------------------------------
|           David Scott #####    ##### 1054 MBNI                     |
|    Doctoral Candidate  #####  #####  205 Zina Pitcher Place        |
|  Neuroscience Program  ############  Ann Arbor, MI 48109-0720      |
|University of Michigan  ###  ##  ###  734.476.2528                  |
------------------------#####    #####-------------------------------- 

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Marsbar-users mailing list
Marsbar-users <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/marsbar-users
Permalink | Reply | 2 comments |
headers
Matthew Brett | 24 Oct 12:03
Picon
Gravatar

Re: Different raw data with one vs. multiple subjects

Hi,

> I have come across a very strange issue using marsbar to extract raw data
> from image series. If I extract raw data (as described in the FAQ) from an
> ROI when applied to images from several different subjects, I get one set of
> values for each image. However if I apply this ROI in the same manner to
> these subjects one at a time, I get a very different set of values from
> these images. I am not using any scaling or anything like that so I don't
> see why these "raw" values would be changing depending on how many of the
> images I extract from. Has anyone come across this, or could offer some
> advice?

Er - I think the only way you could get different values for the
images is if the scaling applied was different.  For example, if you
used the SPM design to extract your images with, SPM has already
applied scaling to the images (for fMRI by default this will be so
that every session has a global brain mean of 100 across all the
images in the session).  One way of getting different values is
therefore to use the SPM design for the individual subjects, and not
to use an SPM design for the multi-subject case,

Best,

Matthew

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
Permalink | Reply | 1 comment |
headers
Murphy, Eric | 25 Oct 23:15

Can't get volume information

Hi Matthew,

I am trying to run an analysis on some raw images, and I am having an error script come up in my matlab window saying “Can’t get volume information for (files)”.

I checked the FAQ page, and made sure it wasn’t the same issue as the “Can’t open image file”m which it does not appear to be.  Checking image names in design worked fine, but the problem persisted.  Any idea what might be wrong with either my files or me?

Thanks.

 

Eric Murphy, M.A.

Research Assistant

Dept. of Psychiatry & Behavioral Neuroscience

Wayne StateUniversityMedicalSchool

emurph <at> med.wayne.edu

(313) 577-1838

 

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Marsbar-users mailing list
Marsbar-users <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/marsbar-users
Permalink | Reply | 1 comment |
headers
Matthew Brett | 26 Oct 15:08
Picon
Gravatar

Re: Can't get volume information

Hi,

> I am trying to run an analysis on some raw images, and I am having an error
> script come up in my matlab window saying "Can't get volume information for
> (files)".

No, not sure.  Have you checked the files haven't been damaged?  Can
you display them in SPM?  Open them in another package like MRIcro?

Best,

Matthew

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
Permalink | Reply | 0 comments |
headers
Matthew Brett | 27 Oct 17:28
Picon
Gravatar

Re: Different raw data with one vs. multiple subjects

Hi,

> > I have come across a very strange issue using marsbar to extract raw data
> > from image series. If I extract raw data (as described in the FAQ) from an
> > ROI when applied to images from several different subjects, I get one set of
> > values for each image. However if I apply this ROI in the same manner to
> > these subjects one at a time, I get a very different set of values from
> > these images. I am not using any scaling or anything like that so I don't
> > see why these "raw" values would be changing depending on how many of the
> > images I extract from. Has anyone come across this, or could offer some
> > advice?

Ah - after having a look at your data I think I understand.  For each
subject there are different numbers of voxels with no data (NaN)
within the ROI.  When marsbar extracts the data, it removes any
columns from the ROI that contain any NaN value.  So, if there is an
NaN value in the ROI for subject2, then, when you extract subject 1's
data with subject 2, the voxel will not contribute to either subject1
or subject2.  If you want all the voxels to be shared voxels in the
ROI across subjects, extract the data with all subjects together.  If
you want all the available voxels in the ROI for each subject, and
don't care that some of the voxels that do contribute to subject 1's
mean will not contribute to subject 2's mean, then do the extraction
separately.

Does that make sense?

Matthew

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
Permalink | Reply | 0 comments |

Gmane