Thanks for the reply! I'm fairly certain that I'm using the same model. In Qdec: 2 discrete variables, 1 nuisance variable. In SPSS: a univariate ANOVA with 2 factors and 1 covariate.
I understand the difference between Qdec and SPSS may be because Qdec looks at the entire cortex. Are there any realistic ways to work around this using something like SPSS?
All I really want is to generate a figure with means for each of my groups, so I can understand what's driving the interaction. Any other suggestions about how to do so?
-----Original Message----- From: greve@nmr.mgh.harvard.edu [mailto:greve@nmr.mgh.harvard.edu] Sent: Thursday, April 28, 2011 9:06 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
a couple of suggestions: 1. make sure that QDEC and SPSS are using the identical model. 2. make sure that the values that you are extracting the same as the input to QDEC/glmfit (eg, don't use thickness for one and volume for another).
It is possible for the ROI analysis to give very different results, but it usually requires that you select the voxels based on an unsigned statistic (eg, an F-test). This opens the possibility that positives and negatives in the ROI sum to 0.
doug
Thanks Jesse. This does sound very similar to my issue.
How did you resolve the problem?
Thanks again!
From: jessebledsoe@gmail.com [mailto:jessebledsoe@gmail.com] On Behalf Of Jesse Bledsoe Sent: Wednesday, April 27, 2011 1:19 PM To: Gallen, Courtney (NIH/NIDA) [F]; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
I'm no expert but I recalled a post similar to yours awhile ago with the same question you ask. Doug and Nick offered an answer that may be of help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg15586.html
Jesse On Wed, Apr 27, 2011 at 1:12 PM, Gallen, Courtney (NIH/NIDA) [F] <gallencl@nida.nih.govmailto:gallencl@nida.nih.gov> wrote: Hi Doug
Have a follow up question regarding significance levels in Qdec related to extracting means from ROIs, etc. I'm posting to the entire mailing list in case anyone else has ideas.
Basically, I extracted mean volumes (from my set of participants) of ROIs for which Qdec reports an interaction between two variables. I want to look at these mean volumes in something like SPSS to see what is driving the interaction/generate figures, etc.
I'm having an issue with one such ROI, in that when I extract the mean volumes and look for this interaction in SPSS, it is not significant (p ~ 0.5, so really not significant at all). Is there any way that an ROI could be significant in Qdec (p of 0.05, Monte Carlo sim), but not in SPSS when I extract means for each participant?
Let me know if this isn't clear. Any thoughts would be appreciated! Courtney
Courtney Gallen Post-baccalaureate IRTA Neuroimaging Research Branch National Institute on Drug Abuse (IRP) 251 Bayview Blvd Suite 200 Baltimore, MD 21224 Tel: (443) 740-2631tel:%28443%29%20740-2631
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 1:06 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Those are the values outside (first column) and inside (2nd col) the ROI. You can add "--id 1" to only report the values inside the ROI.
doug
On 4/26/11 1:02 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
It worked! Now, last question. The txt file has two separate values (?averages). What do they each represent?
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:53 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
try "--slabel fsaverage lh ROI-drawn-on-QDEC-brain.label"
On 4/26/11 12:52 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Actually, I hope I'm wrong about what I thought below. If I specify individual subject's label and run:
mri_segstats --in /prot/Mprot/CTQ/freesurferDone/8579/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 8579 lh /prot/Mprot/CTQ/freesurferDone/8579/label/lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
I still get the same error about dimension mismatch
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 26, 2011 12:43 PM To: 'Douglas Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
Ah I think I see the problem now.
The label I originally specified is the general label created from the ROI drawing in Qdec. Instead, I should use the individual subject's label (from 'Map Label to Subjects') instead, right?
Thank you!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:38 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
You have to use the subject that matches the input, fsaverage in this case. How did you define your label? If you defined it on subject 123, you can transfer it to fsaverage using mri_label2label. doug
On 4/26/11 12:28 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Not a problem at all!
If I use --in, I get an error saying that there is a dimension mismatch between input volume and seg.
Any other thoughts? Thanks again!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 11:40 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney, sorry for the delay. If more than a few days go by without a response, feel free to repost -- I promise not to get annoyed! Sometimes, things get pushed up and out of my mail browser window, and I often don't get back to them. Sorry!
To answer your question, try using --in where you have --seg. The "segmentation" is actually already specified because you've specified a label.
doug
On 4/26/11 11:03 AM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Hi Doug, Sorry to keep bugging you but any thoughts about my response to your e-mail from 2 weeks ago? I've given the exact command line I'm using with mri_segstats. Thanks again for all your help Courtney
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 12, 2011 10:59 AM To: 'Douglas N Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
That may be the problem. ${sub} is an actual person's directory
E.g., sub# 123--command line would be: mri_segstats --seg /prot/Mprot/CTQ/freesurferDone/123/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 123 lh \ lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
-----Original Message----- From: Douglas N Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 12, 2011 10:55 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
What is your full command-line? Without variables, just something that you could put directly into the terminal. Make sure that you are specifying fsaverage as your subject (can't tell below because your use ${sub}).
doug
Gallen, Courtney (NIH/NIDA) [F] wrote:
Sorry, but I can't seem to figure out how to specify input data. I tried '--in' but get an error that the input vol and seg label are not the same dimensions. I've been looking at the help options for mri_segstats (http://surfer.nmr.mgh.harvard.edu/fswiki/mri_segstats) but don't see it
-----Original Message----- From: Douglas N Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Friday, April 08, 2011 10:38 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
The lh.volume.fwhm10.fsaverage.mgh should be your input. It is not a segmentation. You've already specified the seg with the the --slabel command. doug
Gallen, Courtney (NIH/NIDA) [F] wrote:
> One last question (hopefully). I'm specifying the input volume with > --seg, but am getting an error ('cannot do frame average without > input volume'). Am I missing something obvious? > > The entire command looks something like this: > mri_segstats --seg > /prot/Mprot/CTQ/freesurferDone/${sub}/surf/lh.volume.fwhm10.fsaverage.mgh > --slabel ${sub} lh \ > lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt > > Thanks again > -----Original Message----- > From: Douglas N Greve > [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] > Sent: Tuesday, April 05, 2011 4:21 PM > To: Gallen, Courtney (NIH/NIDA) [F] > Cc: Nick Schmansky; Freesurfer Mailing List > Subject: Re: [Freesurfer] Interpreting results from Qdec > > yes, something like ?h.volume.fwhm10.fsaverage.mgh > doug > > Gallen, Courtney (NIH/NIDA) [F] wrote: > > >> Hi Doug >> >> Thanks for helping me with this! >> >> First time around I did everything as is outlined in 'Define a >> Region of Interest' here: >> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis. >> >> I'll try your suggestion now. The input data would be something >> like ?h.volume.fwhm10.fsaverage.mgh, correct? >> >> Thanks again >> Courtney >> >> -----Original Message----- >> From: Douglas N Greve >> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >> Sent: Thursday, March 31, 2011 5:44 PM >> To: Gallen, Courtney (NIH/NIDA) [F] >> Cc: Nick Schmansky; Freesurfer Mailing List >> Subject: Re: [Freesurfer] Interpreting results from Qdec >> >> Hi Courtney, did you transfer the label to each individual subject >> using >> mri_label2label? A better way to replicate your qdec results is to >> run >> mri_segstats on the input data to qdec (y.mgh usually). Specify >> the >> label with "--slabel subject hemi yourlabel". Also specify >> "--avgwf >> subjectdata.txt". This will create this text file with a list of >> the >> input data for each subject averaged over your label. >> >> doug >> >> Gallen, Courtney (NIH/NIDA) [F] wrote: >> >> >> >>> Hi Nick >>> >>> I'm revisiting this issue and have a few more questions. First, I >>> have a significant interaction between two variables on cortical >>> volume. I drew a label on this ROI and ran mris_anatomical_stats >>> on the label (thank you for that suggestion). >>> >>> For stats in the output table file, is the column 'GrayVol' the >>> same thing as the cortical volume I looked at in Qdec? I'm asking >>> because when I export GrayVol values into SPSS, there is no >>> longer a significant interaction between my variables (and if >>> this ROI is significant after multiple comparisons across the >>> brain, I'd expect the mean volume from the ROI to be very >>> significant in SPSS). >>> >>> Please let me know if my questions aren't clear. Thanks in >>> advance >>> Courtney >>> >>> >>> Courtney Gallen >>> Post-baccalaureate IRTA >>> Neuroimaging Research Branch >>> National Institute on Drug Abuse (IRP) >>> 251 Bayview Blvd >>> Suite 200 >>> Baltimore, MD 21224 >>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>> >>> >>> -----Original Message----- >>> From: Nick Schmansky >>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>> Sent: Thursday, January 13, 2011 3:08 PM >>> To: Gallen, Courtney (NIH/NIDA) [F] >>> Cc: Freesurfer Mailing List >>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>> >>> Courtney, >>> >>> if i understand your question, i think one way to do that is to >>> use qdec >>> to draw a label on the blob of interest, and the run 'map label >>> to >>> subjects', which creates a label file in each subjects label dir, >>> then >>> you can run stats on that with mris_anatomical_stats. >>> >>> (btw, i'm putting this answer on the list, as others may have >>> better >>> ideas.) >>> >>> n. >>> >>> >>> >>> On Mon, 2011-01-03 at 16:38 -0500, Gallen, Courtney (NIH/NIDA) >>> [F] >>> wrote: >>> >>> >>> >>> >>>> Hi Nick >>>> >>>> Thanks for the prompt reply. Sorry if my questions seem a bit >>>> naïve--this is my first time using Freesurfer. >>>> >>>> A follow up question to your answer for question 1. I see that >>>> the group data will be plotted in Qdec for significant blobs, >>>> but say there is a significant interaction between two variables >>>> and it's not entirely apparent what is driving this interaction. >>>> Is there a way to extract the mean of this blob for each >>>> individual (i.e., treat it as an ROI or something similar) to >>>> determine what's driving the significance? >>>> >>>> Thanks in advance! >>>> Courtney >>>> >>>> -----Original Message----- >>>> From: Nick Schmansky >>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>> Sent: Thursday, December 30, 2010 2:44 PM >>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>> Cc: >>>> freesurfer@nmr.mgh.harvard.edumailto:freesurfer@nmr.mgh.harvard.edu >>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>> >>>> ans. 1. - the easiest way is to Ctrl- left mouse click on a >>>> blob, and a >>>> plot of the data at that surface vertex will appear. the group >>>> will be >>>> apparent from that (say, demented group is thinner than >>>> non-demented >>>> group). this data is significance data (log p, so '2' is 0.01), >>>> so mean >>>> and stdev wouldnt apply in that case. the file 'y.mgh' is the >>>> raw data >>>> of all subjects in the analysis, so you could extract mean and >>>> stdev >>>> from that. >>>> >>>> ans. 2. - slide the 'annotation' opacity slider to show the >>>> annotation >>>> data (cortical parcellation). the ctrl left click will also put >>>> the >>>> region info for that vertex on the lower left of the display. >>>> >>>> see also: >>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis >>>> >>>> n. >>>> >>>> On Wed, 2010-12-29 at 10:49 -0500, Gallen, Courtney (NIH/NIDA) >>>> [F] >>>> wrote: >>>> >>>> >>>> >>>> >>>>> Hi, >>>>> >>>>> >>>>> >>>>> I'm currently running analyses in Qdec and have identified >>>>> areas that >>>>> show significant effects in my contrasts. I have two questions >>>>> related >>>>> to interpreting these results. >>>>> >>>>> >>>>> >>>>> 1. How can you interpret the effect in each "blob?" (i.e., >>>>> which >>>>> group has a greater cortical surface area, etc.). Right now, >>>>> all I >>>>> know is that there is an effect in certain areas. Can you >>>>> export this >>>>> data to get means and standard errors? >>>>> >>>>> 2. Is there a way to localize where these effects are? >>>>> E.g., a >>>>> tool that tells you this effect is located in __ area of the >>>>> brain? >>>>> >>>>> >>>>> >>>>> Thanks in advance >>>>> >>>>> Courtney >>>>> >>>>> >>>>> >>>>> Courtney Gallen >>>>> >>>>> Post-baccalaureate IRTA >>>>> >>>>> Neuroimaging Research Branch >>>>> >>>>> National Institute on Drug Abuse (IRP) >>>>> >>>>> 251 Bayview Blvd >>>>> >>>>> Suite 200 >>>>> >>>>> Baltimore, MD 21224 >>>>> >>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>> >>>>> >>>>> _______________________________________________ >>>>> Freesurfer mailing list >>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>> >>>>> >>>>> >>>>> >>>> The information in this e-mail is intended only for the person >>>> to whom it is >>>> addressed. If you believe this e-mail was sent to you in error >>>> and the e-mail >>>> contains patient information, please contact the Partners >>>> Compliance HelpLine at >>>> http://www.partners.org/complianceline . If the e-mail was sent >>>> to you in error >>>> but does not contain patient information, please contact the >>>> sender and properly >>>> dispose of the e-mail. >>>> >>>> >>>> >>>> >>>> >>>> >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >>> >>> >>> >>> >>>
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Jesse C. Bledsoe, M.A. Ph.D. Candidate, Clinical Psychology Consortium for Neurodevelopmental Study Department of Psychology 110E Psychology Building Michigan State University East Lansing, MI 48824 ______________________________________
Note: Information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Gallen, I suggest that you extract the values across subjects at a single vertex in fsaverage space, and then attempt to duplicate FS's uncorrected t- and p-values at that vertex with your SPSS model. Those HAVE to be identical if you are truly using the same model (unless there is an unknown bug related to GLM statistics in FS, which I think it quite unlikely). I vaguely recall myself doing something similar at some point (but using SAS instead of SPSS), and the results were indeed the same. But you really need to understand the underlying model used by both FS and in your case SPSS.
Best, -MH
On Wed, 2011-05-04 at 13:57 -0400, Gallen, Courtney (NIH/NIDA) [F] wrote:
Thanks for the reply! I'm fairly certain that I'm using the same model. In Qdec: 2 discrete variables, 1 nuisance variable. In SPSS: a univariate ANOVA with 2 factors and 1 covariate.
I understand the difference between Qdec and SPSS may be because Qdec looks at the entire cortex. Are there any realistic ways to work around this using something like SPSS?
All I really want is to generate a figure with means for each of my groups, so I can understand what's driving the interaction. Any other suggestions about how to do so?
-----Original Message----- From: greve@nmr.mgh.harvard.edu [mailto:greve@nmr.mgh.harvard.edu] Sent: Thursday, April 28, 2011 9:06 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
a couple of suggestions:
- make sure that QDEC and SPSS are using the identical model.
- make sure that the values that you are extracting the same as the
input to QDEC/glmfit (eg, don't use thickness for one and volume for another).
It is possible for the ROI analysis to give very different results, but it usually requires that you select the voxels based on an unsigned statistic (eg, an F-test). This opens the possibility that positives and negatives in the ROI sum to 0.
doug
Thanks Jesse. This does sound very similar to my issue.
How did you resolve the problem?
Thanks again!
From: jessebledsoe@gmail.com [mailto:jessebledsoe@gmail.com] On Behalf Of Jesse Bledsoe Sent: Wednesday, April 27, 2011 1:19 PM To: Gallen, Courtney (NIH/NIDA) [F]; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
I'm no expert but I recalled a post similar to yours awhile ago with the same question you ask. Doug and Nick offered an answer that may be of help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg15586.html
Jesse On Wed, Apr 27, 2011 at 1:12 PM, Gallen, Courtney (NIH/NIDA) [F] <gallencl@nida.nih.govmailto:gallencl@nida.nih.gov> wrote: Hi Doug
Have a follow up question regarding significance levels in Qdec related to extracting means from ROIs, etc. I'm posting to the entire mailing list in case anyone else has ideas.
Basically, I extracted mean volumes (from my set of participants) of ROIs for which Qdec reports an interaction between two variables. I want to look at these mean volumes in something like SPSS to see what is driving the interaction/generate figures, etc.
I'm having an issue with one such ROI, in that when I extract the mean volumes and look for this interaction in SPSS, it is not significant (p ~ 0.5, so really not significant at all). Is there any way that an ROI could be significant in Qdec (p of 0.05, Monte Carlo sim), but not in SPSS when I extract means for each participant?
Let me know if this isn't clear. Any thoughts would be appreciated! Courtney
Courtney Gallen Post-baccalaureate IRTA Neuroimaging Research Branch National Institute on Drug Abuse (IRP) 251 Bayview Blvd Suite 200 Baltimore, MD 21224 Tel: (443) 740-2631tel:%28443%29%20740-2631
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 1:06 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Those are the values outside (first column) and inside (2nd col) the ROI. You can add "--id 1" to only report the values inside the ROI.
doug
On 4/26/11 1:02 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
It worked! Now, last question. The txt file has two separate values (?averages). What do they each represent?
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:53 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
try "--slabel fsaverage lh ROI-drawn-on-QDEC-brain.label"
On 4/26/11 12:52 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Actually, I hope I'm wrong about what I thought below. If I specify individual subject's label and run:
mri_segstats --in /prot/Mprot/CTQ/freesurferDone/8579/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 8579 lh /prot/Mprot/CTQ/freesurferDone/8579/label/lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
I still get the same error about dimension mismatch
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 26, 2011 12:43 PM To: 'Douglas Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
Ah I think I see the problem now.
The label I originally specified is the general label created from the ROI drawing in Qdec. Instead, I should use the individual subject's label (from 'Map Label to Subjects') instead, right?
Thank you!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:38 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
You have to use the subject that matches the input, fsaverage in this case. How did you define your label? If you defined it on subject 123, you can transfer it to fsaverage using mri_label2label. doug
On 4/26/11 12:28 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Not a problem at all!
If I use --in, I get an error saying that there is a dimension mismatch between input volume and seg.
Any other thoughts? Thanks again!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 11:40 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney, sorry for the delay. If more than a few days go by without a response, feel free to repost -- I promise not to get annoyed! Sometimes, things get pushed up and out of my mail browser window, and I often don't get back to them. Sorry!
To answer your question, try using --in where you have --seg. The "segmentation" is actually already specified because you've specified a label.
doug
On 4/26/11 11:03 AM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Hi Doug, Sorry to keep bugging you but any thoughts about my response to your e-mail from 2 weeks ago? I've given the exact command line I'm using with mri_segstats. Thanks again for all your help Courtney
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 12, 2011 10:59 AM To: 'Douglas N Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
That may be the problem. ${sub} is an actual person's directory
E.g., sub# 123--command line would be: mri_segstats --seg /prot/Mprot/CTQ/freesurferDone/123/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 123 lh \ lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
-----Original Message----- From: Douglas N Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 12, 2011 10:55 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
What is your full command-line? Without variables, just something that you could put directly into the terminal. Make sure that you are specifying fsaverage as your subject (can't tell below because your use ${sub}).
doug
Gallen, Courtney (NIH/NIDA) [F] wrote: > Sorry, but I can't seem to figure out how to specify input data. I > tried '--in' but get an error that the input vol and seg label are > not the same dimensions. I've been looking at the help options for > mri_segstats (http://surfer.nmr.mgh.harvard.edu/fswiki/mri_segstats) > but don't see it > > -----Original Message----- > From: Douglas N Greve > [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] > Sent: Friday, April 08, 2011 10:38 AM > To: Gallen, Courtney (NIH/NIDA) [F] > Subject: Re: [Freesurfer] Interpreting results from Qdec > > The lh.volume.fwhm10.fsaverage.mgh should be your input. It is not > a > segmentation. You've already specified the seg with the the --slabel > command. > doug > > > Gallen, Courtney (NIH/NIDA) [F] wrote: > >> One last question (hopefully). I'm specifying the input volume with >> --seg, but am getting an error ('cannot do frame average without >> input volume'). Am I missing something obvious? >> >> The entire command looks something like this: >> mri_segstats --seg >> /prot/Mprot/CTQ/freesurferDone/${sub}/surf/lh.volume.fwhm10.fsaverage.mgh >> --slabel ${sub} lh \ >> lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt >> >> Thanks again >> -----Original Message----- >> From: Douglas N Greve >> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >> Sent: Tuesday, April 05, 2011 4:21 PM >> To: Gallen, Courtney (NIH/NIDA) [F] >> Cc: Nick Schmansky; Freesurfer Mailing List >> Subject: Re: [Freesurfer] Interpreting results from Qdec >> >> yes, something like ?h.volume.fwhm10.fsaverage.mgh >> doug >> >> Gallen, Courtney (NIH/NIDA) [F] wrote: >> >> >>> Hi Doug >>> >>> Thanks for helping me with this! >>> >>> First time around I did everything as is outlined in 'Define a >>> Region of Interest' here: >>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis. >>> >>> I'll try your suggestion now. The input data would be something >>> like ?h.volume.fwhm10.fsaverage.mgh, correct? >>> >>> Thanks again >>> Courtney >>> >>> -----Original Message----- >>> From: Douglas N Greve >>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>> Sent: Thursday, March 31, 2011 5:44 PM >>> To: Gallen, Courtney (NIH/NIDA) [F] >>> Cc: Nick Schmansky; Freesurfer Mailing List >>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>> >>> Hi Courtney, did you transfer the label to each individual subject >>> using >>> mri_label2label? A better way to replicate your qdec results is to >>> run >>> mri_segstats on the input data to qdec (y.mgh usually). Specify >>> the >>> label with "--slabel subject hemi yourlabel". Also specify >>> "--avgwf >>> subjectdata.txt". This will create this text file with a list of >>> the >>> input data for each subject averaged over your label. >>> >>> doug >>> >>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>> >>> >>> >>>> Hi Nick >>>> >>>> I'm revisiting this issue and have a few more questions. First, I >>>> have a significant interaction between two variables on cortical >>>> volume. I drew a label on this ROI and ran mris_anatomical_stats >>>> on the label (thank you for that suggestion). >>>> >>>> For stats in the output table file, is the column 'GrayVol' the >>>> same thing as the cortical volume I looked at in Qdec? I'm asking >>>> because when I export GrayVol values into SPSS, there is no >>>> longer a significant interaction between my variables (and if >>>> this ROI is significant after multiple comparisons across the >>>> brain, I'd expect the mean volume from the ROI to be very >>>> significant in SPSS). >>>> >>>> Please let me know if my questions aren't clear. Thanks in >>>> advance >>>> Courtney >>>> >>>> >>>> Courtney Gallen >>>> Post-baccalaureate IRTA >>>> Neuroimaging Research Branch >>>> National Institute on Drug Abuse (IRP) >>>> 251 Bayview Blvd >>>> Suite 200 >>>> Baltimore, MD 21224 >>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>> >>>> >>>> -----Original Message----- >>>> From: Nick Schmansky >>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>> Sent: Thursday, January 13, 2011 3:08 PM >>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>> Cc: Freesurfer Mailing List >>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>> >>>> Courtney, >>>> >>>> if i understand your question, i think one way to do that is to >>>> use qdec >>>> to draw a label on the blob of interest, and the run 'map label >>>> to >>>> subjects', which creates a label file in each subjects label dir, >>>> then >>>> you can run stats on that with mris_anatomical_stats. >>>> >>>> (btw, i'm putting this answer on the list, as others may have >>>> better >>>> ideas.) >>>> >>>> n. >>>> >>>> >>>> >>>> On Mon, 2011-01-03 at 16:38 -0500, Gallen, Courtney (NIH/NIDA) >>>> [F] >>>> wrote: >>>> >>>> >>>> >>>> >>>>> Hi Nick >>>>> >>>>> Thanks for the prompt reply. Sorry if my questions seem a bit >>>>> naïve--this is my first time using Freesurfer. >>>>> >>>>> A follow up question to your answer for question 1. I see that >>>>> the group data will be plotted in Qdec for significant blobs, >>>>> but say there is a significant interaction between two variables >>>>> and it's not entirely apparent what is driving this interaction. >>>>> Is there a way to extract the mean of this blob for each >>>>> individual (i.e., treat it as an ROI or something similar) to >>>>> determine what's driving the significance? >>>>> >>>>> Thanks in advance! >>>>> Courtney >>>>> >>>>> -----Original Message----- >>>>> From: Nick Schmansky >>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>> Sent: Thursday, December 30, 2010 2:44 PM >>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>> Cc: >>>>> freesurfer@nmr.mgh.harvard.edumailto:freesurfer@nmr.mgh.harvard.edu >>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>> >>>>> ans. 1. - the easiest way is to Ctrl- left mouse click on a >>>>> blob, and a >>>>> plot of the data at that surface vertex will appear. the group >>>>> will be >>>>> apparent from that (say, demented group is thinner than >>>>> non-demented >>>>> group). this data is significance data (log p, so '2' is 0.01), >>>>> so mean >>>>> and stdev wouldnt apply in that case. the file 'y.mgh' is the >>>>> raw data >>>>> of all subjects in the analysis, so you could extract mean and >>>>> stdev >>>>> from that. >>>>> >>>>> ans. 2. - slide the 'annotation' opacity slider to show the >>>>> annotation >>>>> data (cortical parcellation). the ctrl left click will also put >>>>> the >>>>> region info for that vertex on the lower left of the display. >>>>> >>>>> see also: >>>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis >>>>> >>>>> n. >>>>> >>>>> On Wed, 2010-12-29 at 10:49 -0500, Gallen, Courtney (NIH/NIDA) >>>>> [F] >>>>> wrote: >>>>> >>>>> >>>>> >>>>> >>>>>> Hi, >>>>>> >>>>>> >>>>>> >>>>>> I'm currently running analyses in Qdec and have identified >>>>>> areas that >>>>>> show significant effects in my contrasts. I have two questions >>>>>> related >>>>>> to interpreting these results. >>>>>> >>>>>> >>>>>> >>>>>> 1. How can you interpret the effect in each "blob?" (i.e., >>>>>> which >>>>>> group has a greater cortical surface area, etc.). Right now, >>>>>> all I >>>>>> know is that there is an effect in certain areas. Can you >>>>>> export this >>>>>> data to get means and standard errors? >>>>>> >>>>>> 2. Is there a way to localize where these effects are? >>>>>> E.g., a >>>>>> tool that tells you this effect is located in __ area of the >>>>>> brain? >>>>>> >>>>>> >>>>>> >>>>>> Thanks in advance >>>>>> >>>>>> Courtney >>>>>> >>>>>> >>>>>> >>>>>> Courtney Gallen >>>>>> >>>>>> Post-baccalaureate IRTA >>>>>> >>>>>> Neuroimaging Research Branch >>>>>> >>>>>> National Institute on Drug Abuse (IRP) >>>>>> >>>>>> 251 Bayview Blvd >>>>>> >>>>>> Suite 200 >>>>>> >>>>>> Baltimore, MD 21224 >>>>>> >>>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>>> >>>>>> >>>>>> _______________________________________________ >>>>>> Freesurfer mailing list >>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>> >>>>>> >>>>>> >>>>>> >>>>> The information in this e-mail is intended only for the person >>>>> to whom it is >>>>> addressed. If you believe this e-mail was sent to you in error >>>>> and the e-mail >>>>> contains patient information, please contact the Partners >>>>> Compliance HelpLine at >>>>> http://www.partners.org/complianceline . If the e-mail was sent >>>>> to you in error >>>>> but does not contain patient information, please contact the >>>>> sender and properly >>>>> dispose of the e-mail. >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>> _______________________________________________ >>>> Freesurfer mailing list >>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>> >>>> _______________________________________________ >>>> Freesurfer mailing list >>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>> >>>> >>>> >>>> >>>> >>>>
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Jesse C. Bledsoe, M.A. Ph.D. Candidate, Clinical Psychology Consortium for Neurodevelopmental Study Department of Psychology 110E Psychology Building Michigan State University East Lansing, MI 48824 ______________________________________
Note: Information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Thank you so much for your suggestion!
Doug, is there a way to modify the mri_segstats command from before to specify coordinates of a vertex (say the peak from the ROI in my contrast)?
Thanks!
-----Original Message----- From: Michael Harms [mailto:mharms@conte.wustl.edu] Sent: Wednesday, May 04, 2011 2:41 PM To: Gallen, Courtney (NIH/NIDA) [F] Cc: Doug Greve; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Gallen, I suggest that you extract the values across subjects at a single vertex in fsaverage space, and then attempt to duplicate FS's uncorrected t- and p-values at that vertex with your SPSS model. Those HAVE to be identical if you are truly using the same model (unless there is an unknown bug related to GLM statistics in FS, which I think it quite unlikely). I vaguely recall myself doing something similar at some point (but using SAS instead of SPSS), and the results were indeed the same. But you really need to understand the underlying model used by both FS and in your case SPSS.
Best, -MH
On Wed, 2011-05-04 at 13:57 -0400, Gallen, Courtney (NIH/NIDA) [F] wrote:
Thanks for the reply! I'm fairly certain that I'm using the same model. In Qdec: 2 discrete variables, 1 nuisance variable. In SPSS: a univariate ANOVA with 2 factors and 1 covariate.
I understand the difference between Qdec and SPSS may be because Qdec looks at the entire cortex. Are there any realistic ways to work around this using something like SPSS?
All I really want is to generate a figure with means for each of my groups, so I can understand what's driving the interaction. Any other suggestions about how to do so?
-----Original Message----- From: greve@nmr.mgh.harvard.edu [mailto:greve@nmr.mgh.harvard.edu] Sent: Thursday, April 28, 2011 9:06 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
a couple of suggestions:
- make sure that QDEC and SPSS are using the identical model.
- make sure that the values that you are extracting the same as the
input to QDEC/glmfit (eg, don't use thickness for one and volume for another).
It is possible for the ROI analysis to give very different results, but it usually requires that you select the voxels based on an unsigned statistic (eg, an F-test). This opens the possibility that positives and negatives in the ROI sum to 0.
doug
Thanks Jesse. This does sound very similar to my issue.
How did you resolve the problem?
Thanks again!
From: jessebledsoe@gmail.com [mailto:jessebledsoe@gmail.com] On Behalf Of Jesse Bledsoe Sent: Wednesday, April 27, 2011 1:19 PM To: Gallen, Courtney (NIH/NIDA) [F]; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
I'm no expert but I recalled a post similar to yours awhile ago with the same question you ask. Doug and Nick offered an answer that may be of help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg15586.html
Jesse On Wed, Apr 27, 2011 at 1:12 PM, Gallen, Courtney (NIH/NIDA) [F] <gallencl@nida.nih.govmailto:gallencl@nida.nih.gov> wrote: Hi Doug
Have a follow up question regarding significance levels in Qdec related to extracting means from ROIs, etc. I'm posting to the entire mailing list in case anyone else has ideas.
Basically, I extracted mean volumes (from my set of participants) of ROIs for which Qdec reports an interaction between two variables. I want to look at these mean volumes in something like SPSS to see what is driving the interaction/generate figures, etc.
I'm having an issue with one such ROI, in that when I extract the mean volumes and look for this interaction in SPSS, it is not significant (p ~ 0.5, so really not significant at all). Is there any way that an ROI could be significant in Qdec (p of 0.05, Monte Carlo sim), but not in SPSS when I extract means for each participant?
Let me know if this isn't clear. Any thoughts would be appreciated! Courtney
Courtney Gallen Post-baccalaureate IRTA Neuroimaging Research Branch National Institute on Drug Abuse (IRP) 251 Bayview Blvd Suite 200 Baltimore, MD 21224 Tel: (443) 740-2631tel:%28443%29%20740-2631
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 1:06 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Those are the values outside (first column) and inside (2nd col) the ROI. You can add "--id 1" to only report the values inside the ROI.
doug
On 4/26/11 1:02 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
It worked! Now, last question. The txt file has two separate values (?averages). What do they each represent?
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:53 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
try "--slabel fsaverage lh ROI-drawn-on-QDEC-brain.label"
On 4/26/11 12:52 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Actually, I hope I'm wrong about what I thought below. If I specify individual subject's label and run:
mri_segstats --in /prot/Mprot/CTQ/freesurferDone/8579/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 8579 lh /prot/Mprot/CTQ/freesurferDone/8579/label/lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
I still get the same error about dimension mismatch
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 26, 2011 12:43 PM To: 'Douglas Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
Ah I think I see the problem now.
The label I originally specified is the general label created from the ROI drawing in Qdec. Instead, I should use the individual subject's label (from 'Map Label to Subjects') instead, right?
Thank you!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:38 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
You have to use the subject that matches the input, fsaverage in this case. How did you define your label? If you defined it on subject 123, you can transfer it to fsaverage using mri_label2label. doug
On 4/26/11 12:28 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Not a problem at all!
If I use --in, I get an error saying that there is a dimension mismatch between input volume and seg.
Any other thoughts? Thanks again!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 11:40 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney, sorry for the delay. If more than a few days go by without a response, feel free to repost -- I promise not to get annoyed! Sometimes, things get pushed up and out of my mail browser window, and I often don't get back to them. Sorry!
To answer your question, try using --in where you have --seg. The "segmentation" is actually already specified because you've specified a label.
doug
On 4/26/11 11:03 AM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Hi Doug, Sorry to keep bugging you but any thoughts about my response to your e-mail from 2 weeks ago? I've given the exact command line I'm using with mri_segstats. Thanks again for all your help Courtney
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 12, 2011 10:59 AM To: 'Douglas N Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
That may be the problem. ${sub} is an actual person's directory
E.g., sub# 123--command line would be: mri_segstats --seg /prot/Mprot/CTQ/freesurferDone/123/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 123 lh \ lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
-----Original Message----- From: Douglas N Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 12, 2011 10:55 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
What is your full command-line? Without variables, just something that you could put directly into the terminal. Make sure that you are specifying fsaverage as your subject (can't tell below because your use ${sub}).
doug
Gallen, Courtney (NIH/NIDA) [F] wrote: > Sorry, but I can't seem to figure out how to specify input data. I > tried '--in' but get an error that the input vol and seg label are > not the same dimensions. I've been looking at the help options for > mri_segstats (http://surfer.nmr.mgh.harvard.edu/fswiki/mri_segstats) > but don't see it > > -----Original Message----- > From: Douglas N Greve > [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] > Sent: Friday, April 08, 2011 10:38 AM > To: Gallen, Courtney (NIH/NIDA) [F] > Subject: Re: [Freesurfer] Interpreting results from Qdec > > The lh.volume.fwhm10.fsaverage.mgh should be your input. It is not > a > segmentation. You've already specified the seg with the the --slabel > command. > doug > > > Gallen, Courtney (NIH/NIDA) [F] wrote: > >> One last question (hopefully). I'm specifying the input volume with >> --seg, but am getting an error ('cannot do frame average without >> input volume'). Am I missing something obvious? >> >> The entire command looks something like this: >> mri_segstats --seg >> /prot/Mprot/CTQ/freesurferDone/${sub}/surf/lh.volume.fwhm10.fsaverage.mgh >> --slabel ${sub} lh \ >> lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt >> >> Thanks again >> -----Original Message----- >> From: Douglas N Greve >> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >> Sent: Tuesday, April 05, 2011 4:21 PM >> To: Gallen, Courtney (NIH/NIDA) [F] >> Cc: Nick Schmansky; Freesurfer Mailing List >> Subject: Re: [Freesurfer] Interpreting results from Qdec >> >> yes, something like ?h.volume.fwhm10.fsaverage.mgh >> doug >> >> Gallen, Courtney (NIH/NIDA) [F] wrote: >> >> >>> Hi Doug >>> >>> Thanks for helping me with this! >>> >>> First time around I did everything as is outlined in 'Define a >>> Region of Interest' here: >>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis. >>> >>> I'll try your suggestion now. The input data would be something >>> like ?h.volume.fwhm10.fsaverage.mgh, correct? >>> >>> Thanks again >>> Courtney >>> >>> -----Original Message----- >>> From: Douglas N Greve >>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>> Sent: Thursday, March 31, 2011 5:44 PM >>> To: Gallen, Courtney (NIH/NIDA) [F] >>> Cc: Nick Schmansky; Freesurfer Mailing List >>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>> >>> Hi Courtney, did you transfer the label to each individual subject >>> using >>> mri_label2label? A better way to replicate your qdec results is to >>> run >>> mri_segstats on the input data to qdec (y.mgh usually). Specify >>> the >>> label with "--slabel subject hemi yourlabel". Also specify >>> "--avgwf >>> subjectdata.txt". This will create this text file with a list of >>> the >>> input data for each subject averaged over your label. >>> >>> doug >>> >>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>> >>> >>> >>>> Hi Nick >>>> >>>> I'm revisiting this issue and have a few more questions. First, I >>>> have a significant interaction between two variables on cortical >>>> volume. I drew a label on this ROI and ran mris_anatomical_stats >>>> on the label (thank you for that suggestion). >>>> >>>> For stats in the output table file, is the column 'GrayVol' the >>>> same thing as the cortical volume I looked at in Qdec? I'm asking >>>> because when I export GrayVol values into SPSS, there is no >>>> longer a significant interaction between my variables (and if >>>> this ROI is significant after multiple comparisons across the >>>> brain, I'd expect the mean volume from the ROI to be very >>>> significant in SPSS). >>>> >>>> Please let me know if my questions aren't clear. Thanks in >>>> advance >>>> Courtney >>>> >>>> >>>> Courtney Gallen >>>> Post-baccalaureate IRTA >>>> Neuroimaging Research Branch >>>> National Institute on Drug Abuse (IRP) >>>> 251 Bayview Blvd >>>> Suite 200 >>>> Baltimore, MD 21224 >>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>> >>>> >>>> -----Original Message----- >>>> From: Nick Schmansky >>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>> Sent: Thursday, January 13, 2011 3:08 PM >>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>> Cc: Freesurfer Mailing List >>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>> >>>> Courtney, >>>> >>>> if i understand your question, i think one way to do that is to >>>> use qdec >>>> to draw a label on the blob of interest, and the run 'map label >>>> to >>>> subjects', which creates a label file in each subjects label dir, >>>> then >>>> you can run stats on that with mris_anatomical_stats. >>>> >>>> (btw, i'm putting this answer on the list, as others may have >>>> better >>>> ideas.) >>>> >>>> n. >>>> >>>> >>>> >>>> On Mon, 2011-01-03 at 16:38 -0500, Gallen, Courtney (NIH/NIDA) >>>> [F] >>>> wrote: >>>> >>>> >>>> >>>> >>>>> Hi Nick >>>>> >>>>> Thanks for the prompt reply. Sorry if my questions seem a bit >>>>> naïve--this is my first time using Freesurfer. >>>>> >>>>> A follow up question to your answer for question 1. I see that >>>>> the group data will be plotted in Qdec for significant blobs, >>>>> but say there is a significant interaction between two variables >>>>> and it's not entirely apparent what is driving this interaction. >>>>> Is there a way to extract the mean of this blob for each >>>>> individual (i.e., treat it as an ROI or something similar) to >>>>> determine what's driving the significance? >>>>> >>>>> Thanks in advance! >>>>> Courtney >>>>> >>>>> -----Original Message----- >>>>> From: Nick Schmansky >>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>> Sent: Thursday, December 30, 2010 2:44 PM >>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>> Cc: >>>>> freesurfer@nmr.mgh.harvard.edumailto:freesurfer@nmr.mgh.harvard.edu >>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>> >>>>> ans. 1. - the easiest way is to Ctrl- left mouse click on a >>>>> blob, and a >>>>> plot of the data at that surface vertex will appear. the group >>>>> will be >>>>> apparent from that (say, demented group is thinner than >>>>> non-demented >>>>> group). this data is significance data (log p, so '2' is 0.01), >>>>> so mean >>>>> and stdev wouldnt apply in that case. the file 'y.mgh' is the >>>>> raw data >>>>> of all subjects in the analysis, so you could extract mean and >>>>> stdev >>>>> from that. >>>>> >>>>> ans. 2. - slide the 'annotation' opacity slider to show the >>>>> annotation >>>>> data (cortical parcellation). the ctrl left click will also put >>>>> the >>>>> region info for that vertex on the lower left of the display. >>>>> >>>>> see also: >>>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis >>>>> >>>>> n. >>>>> >>>>> On Wed, 2010-12-29 at 10:49 -0500, Gallen, Courtney (NIH/NIDA) >>>>> [F] >>>>> wrote: >>>>> >>>>> >>>>> >>>>> >>>>>> Hi, >>>>>> >>>>>> >>>>>> >>>>>> I'm currently running analyses in Qdec and have identified >>>>>> areas that >>>>>> show significant effects in my contrasts. I have two questions >>>>>> related >>>>>> to interpreting these results. >>>>>> >>>>>> >>>>>> >>>>>> 1. How can you interpret the effect in each "blob?" (i.e., >>>>>> which >>>>>> group has a greater cortical surface area, etc.). Right now, >>>>>> all I >>>>>> know is that there is an effect in certain areas. Can you >>>>>> export this >>>>>> data to get means and standard errors? >>>>>> >>>>>> 2. Is there a way to localize where these effects are? >>>>>> E.g., a >>>>>> tool that tells you this effect is located in __ area of the >>>>>> brain? >>>>>> >>>>>> >>>>>> >>>>>> Thanks in advance >>>>>> >>>>>> Courtney >>>>>> >>>>>> >>>>>> >>>>>> Courtney Gallen >>>>>> >>>>>> Post-baccalaureate IRTA >>>>>> >>>>>> Neuroimaging Research Branch >>>>>> >>>>>> National Institute on Drug Abuse (IRP) >>>>>> >>>>>> 251 Bayview Blvd >>>>>> >>>>>> Suite 200 >>>>>> >>>>>> Baltimore, MD 21224 >>>>>> >>>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>>> >>>>>> >>>>>> _______________________________________________ >>>>>> Freesurfer mailing list >>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>> >>>>>> >>>>>> >>>>>> >>>>> The information in this e-mail is intended only for the person >>>>> to whom it is >>>>> addressed. If you believe this e-mail was sent to you in error >>>>> and the e-mail >>>>> contains patient information, please contact the Partners >>>>> Compliance HelpLine at >>>>> http://www.partners.org/complianceline . If the e-mail was sent >>>>> to you in error >>>>> but does not contain patient information, please contact the >>>>> sender and properly >>>>> dispose of the e-mail. >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>> _______________________________________________ >>>> Freesurfer mailing list >>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>> >>>> _______________________________________________ >>>> Freesurfer mailing list >>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>> >>>> >>>> >>>> >>>> >>>>
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Jesse C. Bledsoe, M.A. Ph.D. Candidate, Clinical Psychology Consortium for Neurodevelopmental Study Department of Psychology 110E Psychology Building Michigan State University East Lansing, MI 48824 ______________________________________
Note: Information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
You can edit the label by hand to only include one vertex. If you got the label from a call to mri_surfcluster, then surfcluster will output a summary file with the max vertex in each cluster.
doug
Gallen, Courtney (NIH/NIDA) [F] wrote:
Thank you so much for your suggestion!
Doug, is there a way to modify the mri_segstats command from before to specify coordinates of a vertex (say the peak from the ROI in my contrast)?
Thanks!
-----Original Message----- From: Michael Harms [mailto:mharms@conte.wustl.edu] Sent: Wednesday, May 04, 2011 2:41 PM To: Gallen, Courtney (NIH/NIDA) [F] Cc: Doug Greve; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Gallen, I suggest that you extract the values across subjects at a single vertex in fsaverage space, and then attempt to duplicate FS's uncorrected t- and p-values at that vertex with your SPSS model. Those HAVE to be identical if you are truly using the same model (unless there is an unknown bug related to GLM statistics in FS, which I think it quite unlikely). I vaguely recall myself doing something similar at some point (but using SAS instead of SPSS), and the results were indeed the same. But you really need to understand the underlying model used by both FS and in your case SPSS.
Best, -MH
On Wed, 2011-05-04 at 13:57 -0400, Gallen, Courtney (NIH/NIDA) [F] wrote:
Thanks for the reply! I'm fairly certain that I'm using the same model. In Qdec: 2 discrete variables, 1 nuisance variable. In SPSS: a univariate ANOVA with 2 factors and 1 covariate.
I understand the difference between Qdec and SPSS may be because Qdec looks at the entire cortex. Are there any realistic ways to work around this using something like SPSS?
All I really want is to generate a figure with means for each of my groups, so I can understand what's driving the interaction. Any other suggestions about how to do so?
-----Original Message----- From: greve@nmr.mgh.harvard.edu [mailto:greve@nmr.mgh.harvard.edu] Sent: Thursday, April 28, 2011 9:06 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
a couple of suggestions:
- make sure that QDEC and SPSS are using the identical model.
- make sure that the values that you are extracting the same as the
input to QDEC/glmfit (eg, don't use thickness for one and volume for another).
It is possible for the ROI analysis to give very different results, but it usually requires that you select the voxels based on an unsigned statistic (eg, an F-test). This opens the possibility that positives and negatives in the ROI sum to 0.
doug
Thanks Jesse. This does sound very similar to my issue.
How did you resolve the problem?
Thanks again!
From: jessebledsoe@gmail.com [mailto:jessebledsoe@gmail.com] On Behalf Of Jesse Bledsoe Sent: Wednesday, April 27, 2011 1:19 PM To: Gallen, Courtney (NIH/NIDA) [F]; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
I'm no expert but I recalled a post similar to yours awhile ago with the same question you ask. Doug and Nick offered an answer that may be of help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg15586.html
Jesse On Wed, Apr 27, 2011 at 1:12 PM, Gallen, Courtney (NIH/NIDA) [F] <gallencl@nida.nih.govmailto:gallencl@nida.nih.gov> wrote: Hi Doug
Have a follow up question regarding significance levels in Qdec related to extracting means from ROIs, etc. I'm posting to the entire mailing list in case anyone else has ideas.
Basically, I extracted mean volumes (from my set of participants) of ROIs for which Qdec reports an interaction between two variables. I want to look at these mean volumes in something like SPSS to see what is driving the interaction/generate figures, etc.
I'm having an issue with one such ROI, in that when I extract the mean volumes and look for this interaction in SPSS, it is not significant (p ~ 0.5, so really not significant at all). Is there any way that an ROI could be significant in Qdec (p of 0.05, Monte Carlo sim), but not in SPSS when I extract means for each participant?
Let me know if this isn't clear. Any thoughts would be appreciated! Courtney
Courtney Gallen Post-baccalaureate IRTA Neuroimaging Research Branch National Institute on Drug Abuse (IRP) 251 Bayview Blvd Suite 200 Baltimore, MD 21224 Tel: (443) 740-2631tel:%28443%29%20740-2631
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 1:06 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Those are the values outside (first column) and inside (2nd col) the ROI. You can add "--id 1" to only report the values inside the ROI.
doug
On 4/26/11 1:02 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
It worked! Now, last question. The txt file has two separate values (?averages). What do they each represent?
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:53 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
try "--slabel fsaverage lh ROI-drawn-on-QDEC-brain.label"
On 4/26/11 12:52 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Actually, I hope I'm wrong about what I thought below. If I specify individual subject's label and run:
mri_segstats --in /prot/Mprot/CTQ/freesurferDone/8579/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 8579 lh /prot/Mprot/CTQ/freesurferDone/8579/label/lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
I still get the same error about dimension mismatch
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 26, 2011 12:43 PM To: 'Douglas Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
Ah I think I see the problem now.
The label I originally specified is the general label created from the ROI drawing in Qdec. Instead, I should use the individual subject's label (from 'Map Label to Subjects') instead, right?
Thank you!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:38 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
You have to use the subject that matches the input, fsaverage in this case. How did you define your label? If you defined it on subject 123, you can transfer it to fsaverage using mri_label2label. doug
On 4/26/11 12:28 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Not a problem at all!
If I use --in, I get an error saying that there is a dimension mismatch between input volume and seg.
Any other thoughts? Thanks again!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 11:40 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney, sorry for the delay. If more than a few days go by without a response, feel free to repost -- I promise not to get annoyed! Sometimes, things get pushed up and out of my mail browser window, and I often don't get back to them. Sorry!
To answer your question, try using --in where you have --seg. The "segmentation" is actually already specified because you've specified a label.
doug
On 4/26/11 11:03 AM, Gallen, Courtney (NIH/NIDA) [F] wrote:
> Hi Doug, > Sorry to keep bugging you but any thoughts about my response to your > e-mail from 2 weeks ago? I've given the exact command line I'm using > with mri_segstats. > Thanks again for all your help > Courtney > > -----Original Message----- > From: Gallen, Courtney (NIH/NIDA) [F] > Sent: Tuesday, April 12, 2011 10:59 AM > To: 'Douglas N Greve' > Subject: RE: [Freesurfer] Interpreting results from Qdec > > That may be the problem. ${sub} is an actual person's directory > > E.g., sub# 123--command line would be: > mri_segstats --seg > /prot/Mprot/CTQ/freesurferDone/123/surf/lh.volume.fwhm10.fsaverage.mgh > --slabel 123 lh \ > lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt > > > > -----Original Message----- > From: Douglas N Greve > [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] > Sent: Tuesday, April 12, 2011 10:55 AM > To: Gallen, Courtney (NIH/NIDA) [F] > Subject: Re: [Freesurfer] Interpreting results from Qdec > > What is your full command-line? Without variables, just something > that > you could put directly into the terminal. Make sure that you are > specifying fsaverage as your subject (can't tell below because your > use > ${sub}). > > doug > > > Gallen, Courtney (NIH/NIDA) [F] wrote: > >> Sorry, but I can't seem to figure out how to specify input data. I >> tried '--in' but get an error that the input vol and seg label are >> not the same dimensions. I've been looking at the help options for >> mri_segstats (http://surfer.nmr.mgh.harvard.edu/fswiki/mri_segstats) >> but don't see it >> >> -----Original Message----- >> From: Douglas N Greve >> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >> Sent: Friday, April 08, 2011 10:38 AM >> To: Gallen, Courtney (NIH/NIDA) [F] >> Subject: Re: [Freesurfer] Interpreting results from Qdec >> >> The lh.volume.fwhm10.fsaverage.mgh should be your input. It is not >> a >> segmentation. You've already specified the seg with the the --slabel >> command. >> doug >> >> >> Gallen, Courtney (NIH/NIDA) [F] wrote: >> >> >>> One last question (hopefully). I'm specifying the input volume with >>> --seg, but am getting an error ('cannot do frame average without >>> input volume'). Am I missing something obvious? >>> >>> The entire command looks something like this: >>> mri_segstats --seg >>> /prot/Mprot/CTQ/freesurferDone/${sub}/surf/lh.volume.fwhm10.fsaverage.mgh >>> --slabel ${sub} lh \ >>> lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt >>> >>> Thanks again >>> -----Original Message----- >>> From: Douglas N Greve >>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>> Sent: Tuesday, April 05, 2011 4:21 PM >>> To: Gallen, Courtney (NIH/NIDA) [F] >>> Cc: Nick Schmansky; Freesurfer Mailing List >>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>> >>> yes, something like ?h.volume.fwhm10.fsaverage.mgh >>> doug >>> >>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>> >>> >>> >>>> Hi Doug >>>> >>>> Thanks for helping me with this! >>>> >>>> First time around I did everything as is outlined in 'Define a >>>> Region of Interest' here: >>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis. >>>> >>>> I'll try your suggestion now. The input data would be something >>>> like ?h.volume.fwhm10.fsaverage.mgh, correct? >>>> >>>> Thanks again >>>> Courtney >>>> >>>> -----Original Message----- >>>> From: Douglas N Greve >>>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>>> Sent: Thursday, March 31, 2011 5:44 PM >>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>> Cc: Nick Schmansky; Freesurfer Mailing List >>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>> >>>> Hi Courtney, did you transfer the label to each individual subject >>>> using >>>> mri_label2label? A better way to replicate your qdec results is to >>>> run >>>> mri_segstats on the input data to qdec (y.mgh usually). Specify >>>> the >>>> label with "--slabel subject hemi yourlabel". Also specify >>>> "--avgwf >>>> subjectdata.txt". This will create this text file with a list of >>>> the >>>> input data for each subject averaged over your label. >>>> >>>> doug >>>> >>>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>>> >>>> >>>> >>>> >>>>> Hi Nick >>>>> >>>>> I'm revisiting this issue and have a few more questions. First, I >>>>> have a significant interaction between two variables on cortical >>>>> volume. I drew a label on this ROI and ran mris_anatomical_stats >>>>> on the label (thank you for that suggestion). >>>>> >>>>> For stats in the output table file, is the column 'GrayVol' the >>>>> same thing as the cortical volume I looked at in Qdec? I'm asking >>>>> because when I export GrayVol values into SPSS, there is no >>>>> longer a significant interaction between my variables (and if >>>>> this ROI is significant after multiple comparisons across the >>>>> brain, I'd expect the mean volume from the ROI to be very >>>>> significant in SPSS). >>>>> >>>>> Please let me know if my questions aren't clear. Thanks in >>>>> advance >>>>> Courtney >>>>> >>>>> >>>>> Courtney Gallen >>>>> Post-baccalaureate IRTA >>>>> Neuroimaging Research Branch >>>>> National Institute on Drug Abuse (IRP) >>>>> 251 Bayview Blvd >>>>> Suite 200 >>>>> Baltimore, MD 21224 >>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>> >>>>> >>>>> -----Original Message----- >>>>> From: Nick Schmansky >>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>> Sent: Thursday, January 13, 2011 3:08 PM >>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>> Cc: Freesurfer Mailing List >>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>> >>>>> Courtney, >>>>> >>>>> if i understand your question, i think one way to do that is to >>>>> use qdec >>>>> to draw a label on the blob of interest, and the run 'map label >>>>> to >>>>> subjects', which creates a label file in each subjects label dir, >>>>> then >>>>> you can run stats on that with mris_anatomical_stats. >>>>> >>>>> (btw, i'm putting this answer on the list, as others may have >>>>> better >>>>> ideas.) >>>>> >>>>> n. >>>>> >>>>> >>>>> >>>>> On Mon, 2011-01-03 at 16:38 -0500, Gallen, Courtney (NIH/NIDA) >>>>> [F] >>>>> wrote: >>>>> >>>>> >>>>> >>>>> >>>>> >>>>>> Hi Nick >>>>>> >>>>>> Thanks for the prompt reply. Sorry if my questions seem a bit >>>>>> naïve--this is my first time using Freesurfer. >>>>>> >>>>>> A follow up question to your answer for question 1. I see that >>>>>> the group data will be plotted in Qdec for significant blobs, >>>>>> but say there is a significant interaction between two variables >>>>>> and it's not entirely apparent what is driving this interaction. >>>>>> Is there a way to extract the mean of this blob for each >>>>>> individual (i.e., treat it as an ROI or something similar) to >>>>>> determine what's driving the significance? >>>>>> >>>>>> Thanks in advance! >>>>>> Courtney >>>>>> >>>>>> -----Original Message----- >>>>>> From: Nick Schmansky >>>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>>> Sent: Thursday, December 30, 2010 2:44 PM >>>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>>> Cc: >>>>>> freesurfer@nmr.mgh.harvard.edumailto:freesurfer@nmr.mgh.harvard.edu >>>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>>> >>>>>> ans. 1. - the easiest way is to Ctrl- left mouse click on a >>>>>> blob, and a >>>>>> plot of the data at that surface vertex will appear. the group >>>>>> will be >>>>>> apparent from that (say, demented group is thinner than >>>>>> non-demented >>>>>> group). this data is significance data (log p, so '2' is 0.01), >>>>>> so mean >>>>>> and stdev wouldnt apply in that case. the file 'y.mgh' is the >>>>>> raw data >>>>>> of all subjects in the analysis, so you could extract mean and >>>>>> stdev >>>>>> from that. >>>>>> >>>>>> ans. 2. - slide the 'annotation' opacity slider to show the >>>>>> annotation >>>>>> data (cortical parcellation). the ctrl left click will also put >>>>>> the >>>>>> region info for that vertex on the lower left of the display. >>>>>> >>>>>> see also: >>>>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis >>>>>> >>>>>> n. >>>>>> >>>>>> On Wed, 2010-12-29 at 10:49 -0500, Gallen, Courtney (NIH/NIDA) >>>>>> [F] >>>>>> wrote: >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>> Hi, >>>>>>> >>>>>>> >>>>>>> >>>>>>> I'm currently running analyses in Qdec and have identified >>>>>>> areas that >>>>>>> show significant effects in my contrasts. I have two questions >>>>>>> related >>>>>>> to interpreting these results. >>>>>>> >>>>>>> >>>>>>> >>>>>>> 1. How can you interpret the effect in each "blob?" (i.e., >>>>>>> which >>>>>>> group has a greater cortical surface area, etc.). Right now, >>>>>>> all I >>>>>>> know is that there is an effect in certain areas. Can you >>>>>>> export this >>>>>>> data to get means and standard errors? >>>>>>> >>>>>>> 2. Is there a way to localize where these effects are? >>>>>>> E.g., a >>>>>>> tool that tells you this effect is located in __ area of the >>>>>>> brain? >>>>>>> >>>>>>> >>>>>>> >>>>>>> Thanks in advance >>>>>>> >>>>>>> Courtney >>>>>>> >>>>>>> >>>>>>> >>>>>>> Courtney Gallen >>>>>>> >>>>>>> Post-baccalaureate IRTA >>>>>>> >>>>>>> Neuroimaging Research Branch >>>>>>> >>>>>>> National Institute on Drug Abuse (IRP) >>>>>>> >>>>>>> 251 Bayview Blvd >>>>>>> >>>>>>> Suite 200 >>>>>>> >>>>>>> Baltimore, MD 21224 >>>>>>> >>>>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>>>> >>>>>>> >>>>>>> _______________________________________________ >>>>>>> Freesurfer mailing list >>>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>> The information in this e-mail is intended only for the person >>>>>> to whom it is >>>>>> addressed. If you believe this e-mail was sent to you in error >>>>>> and the e-mail >>>>>> contains patient information, please contact the Partners >>>>>> Compliance HelpLine at >>>>>> http://www.partners.org/complianceline . If the e-mail was sent >>>>>> to you in error >>>>>> but does not contain patient information, please contact the >>>>>> sender and properly >>>>>> dispose of the e-mail. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> _______________________________________________ >>>>> Freesurfer mailing list >>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>> >>>>> _______________________________________________ >>>>> Freesurfer mailing list >>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>>
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Jesse C. Bledsoe, M.A. Ph.D. Candidate, Clinical Psychology Consortium for Neurodevelopmental Study Department of Psychology 110E Psychology Building Michigan State University East Lansing, MI 48824 ______________________________________
Note: Information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Ok, tried this out and things are still coming out a little bit odd.
I manually edited the label file to just include the peak vertex and then re-ran mri_segstats on that label. When I look at these values in SPSS, things are still not quite significant (p ~ 0.09).
Could this have anything to do with the type of multiple comparison corrections I'm doing in Qdec? Also, is there any documentation on how Qdec runs the model specified (i.e., is it something like a univariate ANOVA at each vertex, or some sort of multivariate analysis?).
Thanks again for your continued help! Courtney
-----Original Message----- From: Douglas N Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Wednesday, May 04, 2011 3:42 PM To: Gallen, Courtney (NIH/NIDA) [F] Cc: Michael Harms; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
You can edit the label by hand to only include one vertex. If you got the label from a call to mri_surfcluster, then surfcluster will output a summary file with the max vertex in each cluster.
doug
Gallen, Courtney (NIH/NIDA) [F] wrote:
Thank you so much for your suggestion!
Doug, is there a way to modify the mri_segstats command from before to specify coordinates of a vertex (say the peak from the ROI in my contrast)?
Thanks!
-----Original Message----- From: Michael Harms [mailto:mharms@conte.wustl.edu] Sent: Wednesday, May 04, 2011 2:41 PM To: Gallen, Courtney (NIH/NIDA) [F] Cc: Doug Greve; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Gallen, I suggest that you extract the values across subjects at a single vertex in fsaverage space, and then attempt to duplicate FS's uncorrected t- and p-values at that vertex with your SPSS model. Those HAVE to be identical if you are truly using the same model (unless there is an unknown bug related to GLM statistics in FS, which I think it quite unlikely). I vaguely recall myself doing something similar at some point (but using SAS instead of SPSS), and the results were indeed the same. But you really need to understand the underlying model used by both FS and in your case SPSS.
Best, -MH
On Wed, 2011-05-04 at 13:57 -0400, Gallen, Courtney (NIH/NIDA) [F] wrote:
Thanks for the reply! I'm fairly certain that I'm using the same model. In Qdec: 2 discrete variables, 1 nuisance variable. In SPSS: a univariate ANOVA with 2 factors and 1 covariate.
I understand the difference between Qdec and SPSS may be because Qdec looks at the entire cortex. Are there any realistic ways to work around this using something like SPSS?
All I really want is to generate a figure with means for each of my groups, so I can understand what's driving the interaction. Any other suggestions about how to do so?
-----Original Message----- From: greve@nmr.mgh.harvard.edu [mailto:greve@nmr.mgh.harvard.edu] Sent: Thursday, April 28, 2011 9:06 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
a couple of suggestions:
- make sure that QDEC and SPSS are using the identical model.
- make sure that the values that you are extracting the same as the
input to QDEC/glmfit (eg, don't use thickness for one and volume for another).
It is possible for the ROI analysis to give very different results, but it usually requires that you select the voxels based on an unsigned statistic (eg, an F-test). This opens the possibility that positives and negatives in the ROI sum to 0.
doug
Thanks Jesse. This does sound very similar to my issue.
How did you resolve the problem?
Thanks again!
From: jessebledsoe@gmail.com [mailto:jessebledsoe@gmail.com] On Behalf Of Jesse Bledsoe Sent: Wednesday, April 27, 2011 1:19 PM To: Gallen, Courtney (NIH/NIDA) [F]; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
I'm no expert but I recalled a post similar to yours awhile ago with the same question you ask. Doug and Nick offered an answer that may be of help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg15586.html
Jesse On Wed, Apr 27, 2011 at 1:12 PM, Gallen, Courtney (NIH/NIDA) [F] <gallencl@nida.nih.govmailto:gallencl@nida.nih.gov> wrote: Hi Doug
Have a follow up question regarding significance levels in Qdec related to extracting means from ROIs, etc. I'm posting to the entire mailing list in case anyone else has ideas.
Basically, I extracted mean volumes (from my set of participants) of ROIs for which Qdec reports an interaction between two variables. I want to look at these mean volumes in something like SPSS to see what is driving the interaction/generate figures, etc.
I'm having an issue with one such ROI, in that when I extract the mean volumes and look for this interaction in SPSS, it is not significant (p ~ 0.5, so really not significant at all). Is there any way that an ROI could be significant in Qdec (p of 0.05, Monte Carlo sim), but not in SPSS when I extract means for each participant?
Let me know if this isn't clear. Any thoughts would be appreciated! Courtney
Courtney Gallen Post-baccalaureate IRTA Neuroimaging Research Branch National Institute on Drug Abuse (IRP) 251 Bayview Blvd Suite 200 Baltimore, MD 21224 Tel: (443) 740-2631tel:%28443%29%20740-2631
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 1:06 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Those are the values outside (first column) and inside (2nd col) the ROI. You can add "--id 1" to only report the values inside the ROI.
doug
On 4/26/11 1:02 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
It worked! Now, last question. The txt file has two separate values (?averages). What do they each represent?
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:53 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
try "--slabel fsaverage lh ROI-drawn-on-QDEC-brain.label"
On 4/26/11 12:52 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Actually, I hope I'm wrong about what I thought below. If I specify individual subject's label and run:
mri_segstats --in /prot/Mprot/CTQ/freesurferDone/8579/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 8579 lh /prot/Mprot/CTQ/freesurferDone/8579/label/lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
I still get the same error about dimension mismatch
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 26, 2011 12:43 PM To: 'Douglas Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
Ah I think I see the problem now.
The label I originally specified is the general label created from the ROI drawing in Qdec. Instead, I should use the individual subject's label (from 'Map Label to Subjects') instead, right?
Thank you!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:38 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
You have to use the subject that matches the input, fsaverage in this case. How did you define your label? If you defined it on subject 123, you can transfer it to fsaverage using mri_label2label. doug
On 4/26/11 12:28 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Not a problem at all!
If I use --in, I get an error saying that there is a dimension mismatch between input volume and seg.
Any other thoughts? Thanks again!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 11:40 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney, sorry for the delay. If more than a few days go by without a response, feel free to repost -- I promise not to get annoyed! Sometimes, things get pushed up and out of my mail browser window, and I often don't get back to them. Sorry!
To answer your question, try using --in where you have --seg. The "segmentation" is actually already specified because you've specified a label.
doug
On 4/26/11 11:03 AM, Gallen, Courtney (NIH/NIDA) [F] wrote:
> Hi Doug, > Sorry to keep bugging you but any thoughts about my response to your > e-mail from 2 weeks ago? I've given the exact command line I'm using > with mri_segstats. > Thanks again for all your help > Courtney > > -----Original Message----- > From: Gallen, Courtney (NIH/NIDA) [F] > Sent: Tuesday, April 12, 2011 10:59 AM > To: 'Douglas N Greve' > Subject: RE: [Freesurfer] Interpreting results from Qdec > > That may be the problem. ${sub} is an actual person's directory > > E.g., sub# 123--command line would be: > mri_segstats --seg > /prot/Mprot/CTQ/freesurferDone/123/surf/lh.volume.fwhm10.fsaverage.mgh > --slabel 123 lh \ > lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt > > > > -----Original Message----- > From: Douglas N Greve > [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] > Sent: Tuesday, April 12, 2011 10:55 AM > To: Gallen, Courtney (NIH/NIDA) [F] > Subject: Re: [Freesurfer] Interpreting results from Qdec > > What is your full command-line? Without variables, just something > that > you could put directly into the terminal. Make sure that you are > specifying fsaverage as your subject (can't tell below because your > use > ${sub}). > > doug > > > Gallen, Courtney (NIH/NIDA) [F] wrote: > >> Sorry, but I can't seem to figure out how to specify input data. I >> tried '--in' but get an error that the input vol and seg label are >> not the same dimensions. I've been looking at the help options for >> mri_segstats (http://surfer.nmr.mgh.harvard.edu/fswiki/mri_segstats) >> but don't see it >> >> -----Original Message----- >> From: Douglas N Greve >> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >> Sent: Friday, April 08, 2011 10:38 AM >> To: Gallen, Courtney (NIH/NIDA) [F] >> Subject: Re: [Freesurfer] Interpreting results from Qdec >> >> The lh.volume.fwhm10.fsaverage.mgh should be your input. It is not >> a >> segmentation. You've already specified the seg with the the --slabel >> command. >> doug >> >> >> Gallen, Courtney (NIH/NIDA) [F] wrote: >> >> >>> One last question (hopefully). I'm specifying the input volume with >>> --seg, but am getting an error ('cannot do frame average without >>> input volume'). Am I missing something obvious? >>> >>> The entire command looks something like this: >>> mri_segstats --seg >>> /prot/Mprot/CTQ/freesurferDone/${sub}/surf/lh.volume.fwhm10.fsaverage.mgh >>> --slabel ${sub} lh \ >>> lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt >>> >>> Thanks again >>> -----Original Message----- >>> From: Douglas N Greve >>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>> Sent: Tuesday, April 05, 2011 4:21 PM >>> To: Gallen, Courtney (NIH/NIDA) [F] >>> Cc: Nick Schmansky; Freesurfer Mailing List >>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>> >>> yes, something like ?h.volume.fwhm10.fsaverage.mgh >>> doug >>> >>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>> >>> >>> >>>> Hi Doug >>>> >>>> Thanks for helping me with this! >>>> >>>> First time around I did everything as is outlined in 'Define a >>>> Region of Interest' here: >>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis. >>>> >>>> I'll try your suggestion now. The input data would be something >>>> like ?h.volume.fwhm10.fsaverage.mgh, correct? >>>> >>>> Thanks again >>>> Courtney >>>> >>>> -----Original Message----- >>>> From: Douglas N Greve >>>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>>> Sent: Thursday, March 31, 2011 5:44 PM >>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>> Cc: Nick Schmansky; Freesurfer Mailing List >>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>> >>>> Hi Courtney, did you transfer the label to each individual subject >>>> using >>>> mri_label2label? A better way to replicate your qdec results is to >>>> run >>>> mri_segstats on the input data to qdec (y.mgh usually). Specify >>>> the >>>> label with "--slabel subject hemi yourlabel". Also specify >>>> "--avgwf >>>> subjectdata.txt". This will create this text file with a list of >>>> the >>>> input data for each subject averaged over your label. >>>> >>>> doug >>>> >>>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>>> >>>> >>>> >>>> >>>>> Hi Nick >>>>> >>>>> I'm revisiting this issue and have a few more questions. First, I >>>>> have a significant interaction between two variables on cortical >>>>> volume. I drew a label on this ROI and ran mris_anatomical_stats >>>>> on the label (thank you for that suggestion). >>>>> >>>>> For stats in the output table file, is the column 'GrayVol' the >>>>> same thing as the cortical volume I looked at in Qdec? I'm asking >>>>> because when I export GrayVol values into SPSS, there is no >>>>> longer a significant interaction between my variables (and if >>>>> this ROI is significant after multiple comparisons across the >>>>> brain, I'd expect the mean volume from the ROI to be very >>>>> significant in SPSS). >>>>> >>>>> Please let me know if my questions aren't clear. Thanks in >>>>> advance >>>>> Courtney >>>>> >>>>> >>>>> Courtney Gallen >>>>> Post-baccalaureate IRTA >>>>> Neuroimaging Research Branch >>>>> National Institute on Drug Abuse (IRP) >>>>> 251 Bayview Blvd >>>>> Suite 200 >>>>> Baltimore, MD 21224 >>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>> >>>>> >>>>> -----Original Message----- >>>>> From: Nick Schmansky >>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>> Sent: Thursday, January 13, 2011 3:08 PM >>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>> Cc: Freesurfer Mailing List >>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>> >>>>> Courtney, >>>>> >>>>> if i understand your question, i think one way to do that is to >>>>> use qdec >>>>> to draw a label on the blob of interest, and the run 'map label >>>>> to >>>>> subjects', which creates a label file in each subjects label dir, >>>>> then >>>>> you can run stats on that with mris_anatomical_stats. >>>>> >>>>> (btw, i'm putting this answer on the list, as others may have >>>>> better >>>>> ideas.) >>>>> >>>>> n. >>>>> >>>>> >>>>> >>>>> On Mon, 2011-01-03 at 16:38 -0500, Gallen, Courtney (NIH/NIDA) >>>>> [F] >>>>> wrote: >>>>> >>>>> >>>>> >>>>> >>>>> >>>>>> Hi Nick >>>>>> >>>>>> Thanks for the prompt reply. Sorry if my questions seem a bit >>>>>> naïve--this is my first time using Freesurfer. >>>>>> >>>>>> A follow up question to your answer for question 1. I see that >>>>>> the group data will be plotted in Qdec for significant blobs, >>>>>> but say there is a significant interaction between two variables >>>>>> and it's not entirely apparent what is driving this interaction. >>>>>> Is there a way to extract the mean of this blob for each >>>>>> individual (i.e., treat it as an ROI or something similar) to >>>>>> determine what's driving the significance? >>>>>> >>>>>> Thanks in advance! >>>>>> Courtney >>>>>> >>>>>> -----Original Message----- >>>>>> From: Nick Schmansky >>>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>>> Sent: Thursday, December 30, 2010 2:44 PM >>>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>>> Cc: >>>>>> freesurfer@nmr.mgh.harvard.edumailto:freesurfer@nmr.mgh.harvard.edu >>>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>>> >>>>>> ans. 1. - the easiest way is to Ctrl- left mouse click on a >>>>>> blob, and a >>>>>> plot of the data at that surface vertex will appear. the group >>>>>> will be >>>>>> apparent from that (say, demented group is thinner than >>>>>> non-demented >>>>>> group). this data is significance data (log p, so '2' is 0.01), >>>>>> so mean >>>>>> and stdev wouldnt apply in that case. the file 'y.mgh' is the >>>>>> raw data >>>>>> of all subjects in the analysis, so you could extract mean and >>>>>> stdev >>>>>> from that. >>>>>> >>>>>> ans. 2. - slide the 'annotation' opacity slider to show the >>>>>> annotation >>>>>> data (cortical parcellation). the ctrl left click will also put >>>>>> the >>>>>> region info for that vertex on the lower left of the display. >>>>>> >>>>>> see also: >>>>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis >>>>>> >>>>>> n. >>>>>> >>>>>> On Wed, 2010-12-29 at 10:49 -0500, Gallen, Courtney (NIH/NIDA) >>>>>> [F] >>>>>> wrote: >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>> Hi, >>>>>>> >>>>>>> >>>>>>> >>>>>>> I'm currently running analyses in Qdec and have identified >>>>>>> areas that >>>>>>> show significant effects in my contrasts. I have two questions >>>>>>> related >>>>>>> to interpreting these results. >>>>>>> >>>>>>> >>>>>>> >>>>>>> 1. How can you interpret the effect in each "blob?" (i.e., >>>>>>> which >>>>>>> group has a greater cortical surface area, etc.). Right now, >>>>>>> all I >>>>>>> know is that there is an effect in certain areas. Can you >>>>>>> export this >>>>>>> data to get means and standard errors? >>>>>>> >>>>>>> 2. Is there a way to localize where these effects are? >>>>>>> E.g., a >>>>>>> tool that tells you this effect is located in __ area of the >>>>>>> brain? >>>>>>> >>>>>>> >>>>>>> >>>>>>> Thanks in advance >>>>>>> >>>>>>> Courtney >>>>>>> >>>>>>> >>>>>>> >>>>>>> Courtney Gallen >>>>>>> >>>>>>> Post-baccalaureate IRTA >>>>>>> >>>>>>> Neuroimaging Research Branch >>>>>>> >>>>>>> National Institute on Drug Abuse (IRP) >>>>>>> >>>>>>> 251 Bayview Blvd >>>>>>> >>>>>>> Suite 200 >>>>>>> >>>>>>> Baltimore, MD 21224 >>>>>>> >>>>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>>>> >>>>>>> >>>>>>> _______________________________________________ >>>>>>> Freesurfer mailing list >>>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>> The information in this e-mail is intended only for the person >>>>>> to whom it is >>>>>> addressed. If you believe this e-mail was sent to you in error >>>>>> and the e-mail >>>>>> contains patient information, please contact the Partners >>>>>> Compliance HelpLine at >>>>>> http://www.partners.org/complianceline . If the e-mail was sent >>>>>> to you in error >>>>>> but does not contain patient information, please contact the >>>>>> sender and properly >>>>>> dispose of the e-mail. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> _______________________________________________ >>>>> Freesurfer mailing list >>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>> >>>>> _______________________________________________ >>>>> Freesurfer mailing list >>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>>
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Jesse C. Bledsoe, M.A. Ph.D. Candidate, Clinical Psychology Consortium for Neurodevelopmental Study Department of Psychology 110E Psychology Building Michigan State University East Lansing, MI 48824 ______________________________________
Note: Information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Hi Courtney, If the t/F stats that you are getting for a single vertex in Qdec do not match what you get in SPSS when analyzing the same input data (up to a trivial precision difference), then the odds are that whatever model you are using in SPSS is not that same as that implemented in Qdec. I'd suggest you check the design matrix (Xg.dat) and associated .mat files for specific contrasts used in Qdec, and use those to figure out if you have the same model in SPSS. If you are not very familiar with how SPSS constructs models (and I'm not) then you might need to get someone involved that knows exactly how SPSS is modeling your data based on your selections.
cheers, -MH
On Thu, 2011-05-12 at 09:22 -0400, Gallen, Courtney (NIH/NIDA) [F] wrote:
Ok, tried this out and things are still coming out a little bit odd.
I manually edited the label file to just include the peak vertex and then re-ran mri_segstats on that label. When I look at these values in SPSS, things are still not quite significant (p ~ 0.09).
Could this have anything to do with the type of multiple comparison corrections I'm doing in Qdec? Also, is there any documentation on how Qdec runs the model specified (i.e., is it something like a univariate ANOVA at each vertex, or some sort of multivariate analysis?).
Thanks again for your continued help! Courtney
-----Original Message----- From: Douglas N Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Wednesday, May 04, 2011 3:42 PM To: Gallen, Courtney (NIH/NIDA) [F] Cc: Michael Harms; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
You can edit the label by hand to only include one vertex. If you got the label from a call to mri_surfcluster, then surfcluster will output a summary file with the max vertex in each cluster.
doug
Gallen, Courtney (NIH/NIDA) [F] wrote:
Thank you so much for your suggestion!
Doug, is there a way to modify the mri_segstats command from before to specify coordinates of a vertex (say the peak from the ROI in my contrast)?
Thanks!
-----Original Message----- From: Michael Harms [mailto:mharms@conte.wustl.edu] Sent: Wednesday, May 04, 2011 2:41 PM To: Gallen, Courtney (NIH/NIDA) [F] Cc: Doug Greve; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Gallen, I suggest that you extract the values across subjects at a single vertex in fsaverage space, and then attempt to duplicate FS's uncorrected t- and p-values at that vertex with your SPSS model. Those HAVE to be identical if you are truly using the same model (unless there is an unknown bug related to GLM statistics in FS, which I think it quite unlikely). I vaguely recall myself doing something similar at some point (but using SAS instead of SPSS), and the results were indeed the same. But you really need to understand the underlying model used by both FS and in your case SPSS.
Best, -MH
On Wed, 2011-05-04 at 13:57 -0400, Gallen, Courtney (NIH/NIDA) [F] wrote:
Thanks for the reply! I'm fairly certain that I'm using the same model. In Qdec: 2 discrete variables, 1 nuisance variable. In SPSS: a univariate ANOVA with 2 factors and 1 covariate.
I understand the difference between Qdec and SPSS may be because Qdec looks at the entire cortex. Are there any realistic ways to work around this using something like SPSS?
All I really want is to generate a figure with means for each of my groups, so I can understand what's driving the interaction. Any other suggestions about how to do so?
-----Original Message----- From: greve@nmr.mgh.harvard.edu [mailto:greve@nmr.mgh.harvard.edu] Sent: Thursday, April 28, 2011 9:06 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
a couple of suggestions:
- make sure that QDEC and SPSS are using the identical model.
- make sure that the values that you are extracting the same as the
input to QDEC/glmfit (eg, don't use thickness for one and volume for another).
It is possible for the ROI analysis to give very different results, but it usually requires that you select the voxels based on an unsigned statistic (eg, an F-test). This opens the possibility that positives and negatives in the ROI sum to 0.
doug
Thanks Jesse. This does sound very similar to my issue.
How did you resolve the problem?
Thanks again!
From: jessebledsoe@gmail.com [mailto:jessebledsoe@gmail.com] On Behalf Of Jesse Bledsoe Sent: Wednesday, April 27, 2011 1:19 PM To: Gallen, Courtney (NIH/NIDA) [F]; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
I'm no expert but I recalled a post similar to yours awhile ago with the same question you ask. Doug and Nick offered an answer that may be of help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg15586.html
Jesse On Wed, Apr 27, 2011 at 1:12 PM, Gallen, Courtney (NIH/NIDA) [F] <gallencl@nida.nih.govmailto:gallencl@nida.nih.gov> wrote: Hi Doug
Have a follow up question regarding significance levels in Qdec related to extracting means from ROIs, etc. I'm posting to the entire mailing list in case anyone else has ideas.
Basically, I extracted mean volumes (from my set of participants) of ROIs for which Qdec reports an interaction between two variables. I want to look at these mean volumes in something like SPSS to see what is driving the interaction/generate figures, etc.
I'm having an issue with one such ROI, in that when I extract the mean volumes and look for this interaction in SPSS, it is not significant (p ~ 0.5, so really not significant at all). Is there any way that an ROI could be significant in Qdec (p of 0.05, Monte Carlo sim), but not in SPSS when I extract means for each participant?
Let me know if this isn't clear. Any thoughts would be appreciated! Courtney
Courtney Gallen Post-baccalaureate IRTA Neuroimaging Research Branch National Institute on Drug Abuse (IRP) 251 Bayview Blvd Suite 200 Baltimore, MD 21224 Tel: (443) 740-2631tel:%28443%29%20740-2631
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 1:06 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Those are the values outside (first column) and inside (2nd col) the ROI. You can add "--id 1" to only report the values inside the ROI.
doug
On 4/26/11 1:02 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
It worked! Now, last question. The txt file has two separate values (?averages). What do they each represent?
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:53 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
try "--slabel fsaverage lh ROI-drawn-on-QDEC-brain.label"
On 4/26/11 12:52 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Actually, I hope I'm wrong about what I thought below. If I specify individual subject's label and run:
mri_segstats --in /prot/Mprot/CTQ/freesurferDone/8579/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 8579 lh /prot/Mprot/CTQ/freesurferDone/8579/label/lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
I still get the same error about dimension mismatch
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 26, 2011 12:43 PM To: 'Douglas Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
Ah I think I see the problem now.
The label I originally specified is the general label created from the ROI drawing in Qdec. Instead, I should use the individual subject's label (from 'Map Label to Subjects') instead, right?
Thank you!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:38 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
You have to use the subject that matches the input, fsaverage in this case. How did you define your label? If you defined it on subject 123, you can transfer it to fsaverage using mri_label2label. doug
On 4/26/11 12:28 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
> Not a problem at all! > > If I use --in, I get an error saying that there is a dimension > mismatch between input volume and seg. > > Any other thoughts? Thanks again! > > -----Original Message----- > From: Douglas Greve > [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] > Sent: Tuesday, April 26, 2011 11:40 AM > To: Gallen, Courtney (NIH/NIDA) [F] > Subject: Re: [Freesurfer] Interpreting results from Qdec > > > Hi Courtney, sorry for the delay. If more than a few days go by > without > a response, feel free to repost -- I promise not to get annoyed! > Sometimes, things get pushed up and out of my mail browser window, and > I > often don't get back to them. Sorry! > > To answer your question, try using --in where you have --seg. The > "segmentation" is actually already specified because you've specified > a > label. > > doug > > On 4/26/11 11:03 AM, Gallen, Courtney (NIH/NIDA) [F] wrote: > >> Hi Doug, >> Sorry to keep bugging you but any thoughts about my response to your >> e-mail from 2 weeks ago? I've given the exact command line I'm using >> with mri_segstats. >> Thanks again for all your help >> Courtney >> >> -----Original Message----- >> From: Gallen, Courtney (NIH/NIDA) [F] >> Sent: Tuesday, April 12, 2011 10:59 AM >> To: 'Douglas N Greve' >> Subject: RE: [Freesurfer] Interpreting results from Qdec >> >> That may be the problem. ${sub} is an actual person's directory >> >> E.g., sub# 123--command line would be: >> mri_segstats --seg >> /prot/Mprot/CTQ/freesurferDone/123/surf/lh.volume.fwhm10.fsaverage.mgh >> --slabel 123 lh \ >> lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt >> >> >> >> -----Original Message----- >> From: Douglas N Greve >> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >> Sent: Tuesday, April 12, 2011 10:55 AM >> To: Gallen, Courtney (NIH/NIDA) [F] >> Subject: Re: [Freesurfer] Interpreting results from Qdec >> >> What is your full command-line? Without variables, just something >> that >> you could put directly into the terminal. Make sure that you are >> specifying fsaverage as your subject (can't tell below because your >> use >> ${sub}). >> >> doug >> >> >> Gallen, Courtney (NIH/NIDA) [F] wrote: >> >>> Sorry, but I can't seem to figure out how to specify input data. I >>> tried '--in' but get an error that the input vol and seg label are >>> not the same dimensions. I've been looking at the help options for >>> mri_segstats (http://surfer.nmr.mgh.harvard.edu/fswiki/mri_segstats) >>> but don't see it >>> >>> -----Original Message----- >>> From: Douglas N Greve >>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>> Sent: Friday, April 08, 2011 10:38 AM >>> To: Gallen, Courtney (NIH/NIDA) [F] >>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>> >>> The lh.volume.fwhm10.fsaverage.mgh should be your input. It is not >>> a >>> segmentation. You've already specified the seg with the the --slabel >>> command. >>> doug >>> >>> >>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>> >>> >>>> One last question (hopefully). I'm specifying the input volume with >>>> --seg, but am getting an error ('cannot do frame average without >>>> input volume'). Am I missing something obvious? >>>> >>>> The entire command looks something like this: >>>> mri_segstats --seg >>>> /prot/Mprot/CTQ/freesurferDone/${sub}/surf/lh.volume.fwhm10.fsaverage.mgh >>>> --slabel ${sub} lh \ >>>> lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt >>>> >>>> Thanks again >>>> -----Original Message----- >>>> From: Douglas N Greve >>>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>>> Sent: Tuesday, April 05, 2011 4:21 PM >>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>> Cc: Nick Schmansky; Freesurfer Mailing List >>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>> >>>> yes, something like ?h.volume.fwhm10.fsaverage.mgh >>>> doug >>>> >>>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>>> >>>> >>>> >>>>> Hi Doug >>>>> >>>>> Thanks for helping me with this! >>>>> >>>>> First time around I did everything as is outlined in 'Define a >>>>> Region of Interest' here: >>>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis. >>>>> >>>>> I'll try your suggestion now. The input data would be something >>>>> like ?h.volume.fwhm10.fsaverage.mgh, correct? >>>>> >>>>> Thanks again >>>>> Courtney >>>>> >>>>> -----Original Message----- >>>>> From: Douglas N Greve >>>>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>>>> Sent: Thursday, March 31, 2011 5:44 PM >>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>> Cc: Nick Schmansky; Freesurfer Mailing List >>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>> >>>>> Hi Courtney, did you transfer the label to each individual subject >>>>> using >>>>> mri_label2label? A better way to replicate your qdec results is to >>>>> run >>>>> mri_segstats on the input data to qdec (y.mgh usually). Specify >>>>> the >>>>> label with "--slabel subject hemi yourlabel". Also specify >>>>> "--avgwf >>>>> subjectdata.txt". This will create this text file with a list of >>>>> the >>>>> input data for each subject averaged over your label. >>>>> >>>>> doug >>>>> >>>>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>>>> >>>>> >>>>> >>>>> >>>>>> Hi Nick >>>>>> >>>>>> I'm revisiting this issue and have a few more questions. First, I >>>>>> have a significant interaction between two variables on cortical >>>>>> volume. I drew a label on this ROI and ran mris_anatomical_stats >>>>>> on the label (thank you for that suggestion). >>>>>> >>>>>> For stats in the output table file, is the column 'GrayVol' the >>>>>> same thing as the cortical volume I looked at in Qdec? I'm asking >>>>>> because when I export GrayVol values into SPSS, there is no >>>>>> longer a significant interaction between my variables (and if >>>>>> this ROI is significant after multiple comparisons across the >>>>>> brain, I'd expect the mean volume from the ROI to be very >>>>>> significant in SPSS). >>>>>> >>>>>> Please let me know if my questions aren't clear. Thanks in >>>>>> advance >>>>>> Courtney >>>>>> >>>>>> >>>>>> Courtney Gallen >>>>>> Post-baccalaureate IRTA >>>>>> Neuroimaging Research Branch >>>>>> National Institute on Drug Abuse (IRP) >>>>>> 251 Bayview Blvd >>>>>> Suite 200 >>>>>> Baltimore, MD 21224 >>>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>>> >>>>>> >>>>>> -----Original Message----- >>>>>> From: Nick Schmansky >>>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>>> Sent: Thursday, January 13, 2011 3:08 PM >>>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>>> Cc: Freesurfer Mailing List >>>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>>> >>>>>> Courtney, >>>>>> >>>>>> if i understand your question, i think one way to do that is to >>>>>> use qdec >>>>>> to draw a label on the blob of interest, and the run 'map label >>>>>> to >>>>>> subjects', which creates a label file in each subjects label dir, >>>>>> then >>>>>> you can run stats on that with mris_anatomical_stats. >>>>>> >>>>>> (btw, i'm putting this answer on the list, as others may have >>>>>> better >>>>>> ideas.) >>>>>> >>>>>> n. >>>>>> >>>>>> >>>>>> >>>>>> On Mon, 2011-01-03 at 16:38 -0500, Gallen, Courtney (NIH/NIDA) >>>>>> [F] >>>>>> wrote: >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>> Hi Nick >>>>>>> >>>>>>> Thanks for the prompt reply. Sorry if my questions seem a bit >>>>>>> naïve--this is my first time using Freesurfer. >>>>>>> >>>>>>> A follow up question to your answer for question 1. I see that >>>>>>> the group data will be plotted in Qdec for significant blobs, >>>>>>> but say there is a significant interaction between two variables >>>>>>> and it's not entirely apparent what is driving this interaction. >>>>>>> Is there a way to extract the mean of this blob for each >>>>>>> individual (i.e., treat it as an ROI or something similar) to >>>>>>> determine what's driving the significance? >>>>>>> >>>>>>> Thanks in advance! >>>>>>> Courtney >>>>>>> >>>>>>> -----Original Message----- >>>>>>> From: Nick Schmansky >>>>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>>>> Sent: Thursday, December 30, 2010 2:44 PM >>>>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>>>> Cc: >>>>>>> freesurfer@nmr.mgh.harvard.edumailto:freesurfer@nmr.mgh.harvard.edu >>>>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>>>> >>>>>>> ans. 1. - the easiest way is to Ctrl- left mouse click on a >>>>>>> blob, and a >>>>>>> plot of the data at that surface vertex will appear. the group >>>>>>> will be >>>>>>> apparent from that (say, demented group is thinner than >>>>>>> non-demented >>>>>>> group). this data is significance data (log p, so '2' is 0.01), >>>>>>> so mean >>>>>>> and stdev wouldnt apply in that case. the file 'y.mgh' is the >>>>>>> raw data >>>>>>> of all subjects in the analysis, so you could extract mean and >>>>>>> stdev >>>>>>> from that. >>>>>>> >>>>>>> ans. 2. - slide the 'annotation' opacity slider to show the >>>>>>> annotation >>>>>>> data (cortical parcellation). the ctrl left click will also put >>>>>>> the >>>>>>> region info for that vertex on the lower left of the display. >>>>>>> >>>>>>> see also: >>>>>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis >>>>>>> >>>>>>> n. >>>>>>> >>>>>>> On Wed, 2010-12-29 at 10:49 -0500, Gallen, Courtney (NIH/NIDA) >>>>>>> [F] >>>>>>> wrote: >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> Hi, >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> I'm currently running analyses in Qdec and have identified >>>>>>>> areas that >>>>>>>> show significant effects in my contrasts. I have two questions >>>>>>>> related >>>>>>>> to interpreting these results. >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> 1. How can you interpret the effect in each "blob?" (i.e., >>>>>>>> which >>>>>>>> group has a greater cortical surface area, etc.). Right now, >>>>>>>> all I >>>>>>>> know is that there is an effect in certain areas. Can you >>>>>>>> export this >>>>>>>> data to get means and standard errors? >>>>>>>> >>>>>>>> 2. Is there a way to localize where these effects are? >>>>>>>> E.g., a >>>>>>>> tool that tells you this effect is located in __ area of the >>>>>>>> brain? >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> Thanks in advance >>>>>>>> >>>>>>>> Courtney >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> Courtney Gallen >>>>>>>> >>>>>>>> Post-baccalaureate IRTA >>>>>>>> >>>>>>>> Neuroimaging Research Branch >>>>>>>> >>>>>>>> National Institute on Drug Abuse (IRP) >>>>>>>> >>>>>>>> 251 Bayview Blvd >>>>>>>> >>>>>>>> Suite 200 >>>>>>>> >>>>>>>> Baltimore, MD 21224 >>>>>>>> >>>>>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>>>>> >>>>>>>> >>>>>>>> _______________________________________________ >>>>>>>> Freesurfer mailing list >>>>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> The information in this e-mail is intended only for the person >>>>>>> to whom it is >>>>>>> addressed. If you believe this e-mail was sent to you in error >>>>>>> and the e-mail >>>>>>> contains patient information, please contact the Partners >>>>>>> Compliance HelpLine at >>>>>>> http://www.partners.org/complianceline . If the e-mail was sent >>>>>>> to you in error >>>>>>> but does not contain patient information, please contact the >>>>>>> sender and properly >>>>>>> dispose of the e-mail. >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>> _______________________________________________ >>>>>> Freesurfer mailing list >>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>> >>>>>> _______________________________________________ >>>>>> Freesurfer mailing list >>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Jesse C. Bledsoe, M.A. Ph.D. Candidate, Clinical Psychology Consortium for Neurodevelopmental Study Department of Psychology 110E Psychology Building Michigan State University East Lansing, MI 48824 ______________________________________
Note: Information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Also, just to make this explicit: I'd compare actual statistic (t or F) values, since comparing p-value requires that you also account for one- sided vs. two-sided statistical testing.
-MH
On Thu, 2011-05-12 at 09:22 -0400, Gallen, Courtney (NIH/NIDA) [F] wrote:
Ok, tried this out and things are still coming out a little bit odd.
I manually edited the label file to just include the peak vertex and then re-ran mri_segstats on that label. When I look at these values in SPSS, things are still not quite significant (p ~ 0.09).
Could this have anything to do with the type of multiple comparison corrections I'm doing in Qdec? Also, is there any documentation on how Qdec runs the model specified (i.e., is it something like a univariate ANOVA at each vertex, or some sort of multivariate analysis?).
Thanks again for your continued help! Courtney
-----Original Message----- From: Douglas N Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Wednesday, May 04, 2011 3:42 PM To: Gallen, Courtney (NIH/NIDA) [F] Cc: Michael Harms; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
You can edit the label by hand to only include one vertex. If you got the label from a call to mri_surfcluster, then surfcluster will output a summary file with the max vertex in each cluster.
doug
Gallen, Courtney (NIH/NIDA) [F] wrote:
Thank you so much for your suggestion!
Doug, is there a way to modify the mri_segstats command from before to specify coordinates of a vertex (say the peak from the ROI in my contrast)?
Thanks!
-----Original Message----- From: Michael Harms [mailto:mharms@conte.wustl.edu] Sent: Wednesday, May 04, 2011 2:41 PM To: Gallen, Courtney (NIH/NIDA) [F] Cc: Doug Greve; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Gallen, I suggest that you extract the values across subjects at a single vertex in fsaverage space, and then attempt to duplicate FS's uncorrected t- and p-values at that vertex with your SPSS model. Those HAVE to be identical if you are truly using the same model (unless there is an unknown bug related to GLM statistics in FS, which I think it quite unlikely). I vaguely recall myself doing something similar at some point (but using SAS instead of SPSS), and the results were indeed the same. But you really need to understand the underlying model used by both FS and in your case SPSS.
Best, -MH
On Wed, 2011-05-04 at 13:57 -0400, Gallen, Courtney (NIH/NIDA) [F] wrote:
Thanks for the reply! I'm fairly certain that I'm using the same model. In Qdec: 2 discrete variables, 1 nuisance variable. In SPSS: a univariate ANOVA with 2 factors and 1 covariate.
I understand the difference between Qdec and SPSS may be because Qdec looks at the entire cortex. Are there any realistic ways to work around this using something like SPSS?
All I really want is to generate a figure with means for each of my groups, so I can understand what's driving the interaction. Any other suggestions about how to do so?
-----Original Message----- From: greve@nmr.mgh.harvard.edu [mailto:greve@nmr.mgh.harvard.edu] Sent: Thursday, April 28, 2011 9:06 AM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
a couple of suggestions:
- make sure that QDEC and SPSS are using the identical model.
- make sure that the values that you are extracting the same as the
input to QDEC/glmfit (eg, don't use thickness for one and volume for another).
It is possible for the ROI analysis to give very different results, but it usually requires that you select the voxels based on an unsigned statistic (eg, an F-test). This opens the possibility that positives and negatives in the ROI sum to 0.
doug
Thanks Jesse. This does sound very similar to my issue.
How did you resolve the problem?
Thanks again!
From: jessebledsoe@gmail.com [mailto:jessebledsoe@gmail.com] On Behalf Of Jesse Bledsoe Sent: Wednesday, April 27, 2011 1:19 PM To: Gallen, Courtney (NIH/NIDA) [F]; Freesurfer Mailing List Subject: Re: [Freesurfer] Interpreting results from Qdec
Hi Courtney,
I'm no expert but I recalled a post similar to yours awhile ago with the same question you ask. Doug and Nick offered an answer that may be of help:
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg15586.html
Jesse On Wed, Apr 27, 2011 at 1:12 PM, Gallen, Courtney (NIH/NIDA) [F] <gallencl@nida.nih.govmailto:gallencl@nida.nih.gov> wrote: Hi Doug
Have a follow up question regarding significance levels in Qdec related to extracting means from ROIs, etc. I'm posting to the entire mailing list in case anyone else has ideas.
Basically, I extracted mean volumes (from my set of participants) of ROIs for which Qdec reports an interaction between two variables. I want to look at these mean volumes in something like SPSS to see what is driving the interaction/generate figures, etc.
I'm having an issue with one such ROI, in that when I extract the mean volumes and look for this interaction in SPSS, it is not significant (p ~ 0.5, so really not significant at all). Is there any way that an ROI could be significant in Qdec (p of 0.05, Monte Carlo sim), but not in SPSS when I extract means for each participant?
Let me know if this isn't clear. Any thoughts would be appreciated! Courtney
Courtney Gallen Post-baccalaureate IRTA Neuroimaging Research Branch National Institute on Drug Abuse (IRP) 251 Bayview Blvd Suite 200 Baltimore, MD 21224 Tel: (443) 740-2631tel:%28443%29%20740-2631
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 1:06 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
Those are the values outside (first column) and inside (2nd col) the ROI. You can add "--id 1" to only report the values inside the ROI.
doug
On 4/26/11 1:02 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
It worked! Now, last question. The txt file has two separate values (?averages). What do they each represent?
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:53 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
try "--slabel fsaverage lh ROI-drawn-on-QDEC-brain.label"
On 4/26/11 12:52 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
Actually, I hope I'm wrong about what I thought below. If I specify individual subject's label and run:
mri_segstats --in /prot/Mprot/CTQ/freesurferDone/8579/surf/lh.volume.fwhm10.fsaverage.mgh --slabel 8579 lh /prot/Mprot/CTQ/freesurferDone/8579/label/lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt
I still get the same error about dimension mismatch
-----Original Message----- From: Gallen, Courtney (NIH/NIDA) [F] Sent: Tuesday, April 26, 2011 12:43 PM To: 'Douglas Greve' Subject: RE: [Freesurfer] Interpreting results from Qdec
Ah I think I see the problem now.
The label I originally specified is the general label created from the ROI drawing in Qdec. Instead, I should use the individual subject's label (from 'Map Label to Subjects') instead, right?
Thank you!
-----Original Message----- From: Douglas Greve [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] Sent: Tuesday, April 26, 2011 12:38 PM To: Gallen, Courtney (NIH/NIDA) [F] Subject: Re: [Freesurfer] Interpreting results from Qdec
You have to use the subject that matches the input, fsaverage in this case. How did you define your label? If you defined it on subject 123, you can transfer it to fsaverage using mri_label2label. doug
On 4/26/11 12:28 PM, Gallen, Courtney (NIH/NIDA) [F] wrote:
> Not a problem at all! > > If I use --in, I get an error saying that there is a dimension > mismatch between input volume and seg. > > Any other thoughts? Thanks again! > > -----Original Message----- > From: Douglas Greve > [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] > Sent: Tuesday, April 26, 2011 11:40 AM > To: Gallen, Courtney (NIH/NIDA) [F] > Subject: Re: [Freesurfer] Interpreting results from Qdec > > > Hi Courtney, sorry for the delay. If more than a few days go by > without > a response, feel free to repost -- I promise not to get annoyed! > Sometimes, things get pushed up and out of my mail browser window, and > I > often don't get back to them. Sorry! > > To answer your question, try using --in where you have --seg. The > "segmentation" is actually already specified because you've specified > a > label. > > doug > > On 4/26/11 11:03 AM, Gallen, Courtney (NIH/NIDA) [F] wrote: > >> Hi Doug, >> Sorry to keep bugging you but any thoughts about my response to your >> e-mail from 2 weeks ago? I've given the exact command line I'm using >> with mri_segstats. >> Thanks again for all your help >> Courtney >> >> -----Original Message----- >> From: Gallen, Courtney (NIH/NIDA) [F] >> Sent: Tuesday, April 12, 2011 10:59 AM >> To: 'Douglas N Greve' >> Subject: RE: [Freesurfer] Interpreting results from Qdec >> >> That may be the problem. ${sub} is an actual person's directory >> >> E.g., sub# 123--command line would be: >> mri_segstats --seg >> /prot/Mprot/CTQ/freesurferDone/123/surf/lh.volume.fwhm10.fsaverage.mgh >> --slabel 123 lh \ >> lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt >> >> >> >> -----Original Message----- >> From: Douglas N Greve >> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >> Sent: Tuesday, April 12, 2011 10:55 AM >> To: Gallen, Courtney (NIH/NIDA) [F] >> Subject: Re: [Freesurfer] Interpreting results from Qdec >> >> What is your full command-line? Without variables, just something >> that >> you could put directly into the terminal. Make sure that you are >> specifying fsaverage as your subject (can't tell below because your >> use >> ${sub}). >> >> doug >> >> >> Gallen, Courtney (NIH/NIDA) [F] wrote: >> >>> Sorry, but I can't seem to figure out how to specify input data. I >>> tried '--in' but get an error that the input vol and seg label are >>> not the same dimensions. I've been looking at the help options for >>> mri_segstats (http://surfer.nmr.mgh.harvard.edu/fswiki/mri_segstats) >>> but don't see it >>> >>> -----Original Message----- >>> From: Douglas N Greve >>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>> Sent: Friday, April 08, 2011 10:38 AM >>> To: Gallen, Courtney (NIH/NIDA) [F] >>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>> >>> The lh.volume.fwhm10.fsaverage.mgh should be your input. It is not >>> a >>> segmentation. You've already specified the seg with the the --slabel >>> command. >>> doug >>> >>> >>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>> >>> >>>> One last question (hopefully). I'm specifying the input volume with >>>> --seg, but am getting an error ('cannot do frame average without >>>> input volume'). Am I missing something obvious? >>>> >>>> The entire command looks something like this: >>>> mri_segstats --seg >>>> /prot/Mprot/CTQ/freesurferDone/${sub}/surf/lh.volume.fwhm10.fsaverage.mgh >>>> --slabel ${sub} lh \ >>>> lh.SupTemp_CTQxBDNFctrls.label --avgwf lhSupTempctrls.txt >>>> >>>> Thanks again >>>> -----Original Message----- >>>> From: Douglas N Greve >>>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>>> Sent: Tuesday, April 05, 2011 4:21 PM >>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>> Cc: Nick Schmansky; Freesurfer Mailing List >>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>> >>>> yes, something like ?h.volume.fwhm10.fsaverage.mgh >>>> doug >>>> >>>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>>> >>>> >>>> >>>>> Hi Doug >>>>> >>>>> Thanks for helping me with this! >>>>> >>>>> First time around I did everything as is outlined in 'Define a >>>>> Region of Interest' here: >>>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis. >>>>> >>>>> I'll try your suggestion now. The input data would be something >>>>> like ?h.volume.fwhm10.fsaverage.mgh, correct? >>>>> >>>>> Thanks again >>>>> Courtney >>>>> >>>>> -----Original Message----- >>>>> From: Douglas N Greve >>>>> [mailto:greve@nmr.mgh.harvard.edumailto:greve@nmr.mgh.harvard.edu] >>>>> Sent: Thursday, March 31, 2011 5:44 PM >>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>> Cc: Nick Schmansky; Freesurfer Mailing List >>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>> >>>>> Hi Courtney, did you transfer the label to each individual subject >>>>> using >>>>> mri_label2label? A better way to replicate your qdec results is to >>>>> run >>>>> mri_segstats on the input data to qdec (y.mgh usually). Specify >>>>> the >>>>> label with "--slabel subject hemi yourlabel". Also specify >>>>> "--avgwf >>>>> subjectdata.txt". This will create this text file with a list of >>>>> the >>>>> input data for each subject averaged over your label. >>>>> >>>>> doug >>>>> >>>>> Gallen, Courtney (NIH/NIDA) [F] wrote: >>>>> >>>>> >>>>> >>>>> >>>>>> Hi Nick >>>>>> >>>>>> I'm revisiting this issue and have a few more questions. First, I >>>>>> have a significant interaction between two variables on cortical >>>>>> volume. I drew a label on this ROI and ran mris_anatomical_stats >>>>>> on the label (thank you for that suggestion). >>>>>> >>>>>> For stats in the output table file, is the column 'GrayVol' the >>>>>> same thing as the cortical volume I looked at in Qdec? I'm asking >>>>>> because when I export GrayVol values into SPSS, there is no >>>>>> longer a significant interaction between my variables (and if >>>>>> this ROI is significant after multiple comparisons across the >>>>>> brain, I'd expect the mean volume from the ROI to be very >>>>>> significant in SPSS). >>>>>> >>>>>> Please let me know if my questions aren't clear. Thanks in >>>>>> advance >>>>>> Courtney >>>>>> >>>>>> >>>>>> Courtney Gallen >>>>>> Post-baccalaureate IRTA >>>>>> Neuroimaging Research Branch >>>>>> National Institute on Drug Abuse (IRP) >>>>>> 251 Bayview Blvd >>>>>> Suite 200 >>>>>> Baltimore, MD 21224 >>>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>>> >>>>>> >>>>>> -----Original Message----- >>>>>> From: Nick Schmansky >>>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>>> Sent: Thursday, January 13, 2011 3:08 PM >>>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>>> Cc: Freesurfer Mailing List >>>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>>> >>>>>> Courtney, >>>>>> >>>>>> if i understand your question, i think one way to do that is to >>>>>> use qdec >>>>>> to draw a label on the blob of interest, and the run 'map label >>>>>> to >>>>>> subjects', which creates a label file in each subjects label dir, >>>>>> then >>>>>> you can run stats on that with mris_anatomical_stats. >>>>>> >>>>>> (btw, i'm putting this answer on the list, as others may have >>>>>> better >>>>>> ideas.) >>>>>> >>>>>> n. >>>>>> >>>>>> >>>>>> >>>>>> On Mon, 2011-01-03 at 16:38 -0500, Gallen, Courtney (NIH/NIDA) >>>>>> [F] >>>>>> wrote: >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>> Hi Nick >>>>>>> >>>>>>> Thanks for the prompt reply. Sorry if my questions seem a bit >>>>>>> naïve--this is my first time using Freesurfer. >>>>>>> >>>>>>> A follow up question to your answer for question 1. I see that >>>>>>> the group data will be plotted in Qdec for significant blobs, >>>>>>> but say there is a significant interaction between two variables >>>>>>> and it's not entirely apparent what is driving this interaction. >>>>>>> Is there a way to extract the mean of this blob for each >>>>>>> individual (i.e., treat it as an ROI or something similar) to >>>>>>> determine what's driving the significance? >>>>>>> >>>>>>> Thanks in advance! >>>>>>> Courtney >>>>>>> >>>>>>> -----Original Message----- >>>>>>> From: Nick Schmansky >>>>>>> [mailto:nicks@nmr.mgh.harvard.edumailto:nicks@nmr.mgh.harvard.edu] >>>>>>> Sent: Thursday, December 30, 2010 2:44 PM >>>>>>> To: Gallen, Courtney (NIH/NIDA) [F] >>>>>>> Cc: >>>>>>> freesurfer@nmr.mgh.harvard.edumailto:freesurfer@nmr.mgh.harvard.edu >>>>>>> Subject: Re: [Freesurfer] Interpreting results from Qdec >>>>>>> >>>>>>> ans. 1. - the easiest way is to Ctrl- left mouse click on a >>>>>>> blob, and a >>>>>>> plot of the data at that surface vertex will appear. the group >>>>>>> will be >>>>>>> apparent from that (say, demented group is thinner than >>>>>>> non-demented >>>>>>> group). this data is significance data (log p, so '2' is 0.01), >>>>>>> so mean >>>>>>> and stdev wouldnt apply in that case. the file 'y.mgh' is the >>>>>>> raw data >>>>>>> of all subjects in the analysis, so you could extract mean and >>>>>>> stdev >>>>>>> from that. >>>>>>> >>>>>>> ans. 2. - slide the 'annotation' opacity slider to show the >>>>>>> annotation >>>>>>> data (cortical parcellation). the ctrl left click will also put >>>>>>> the >>>>>>> region info for that vertex on the lower left of the display. >>>>>>> >>>>>>> see also: >>>>>>> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis >>>>>>> >>>>>>> n. >>>>>>> >>>>>>> On Wed, 2010-12-29 at 10:49 -0500, Gallen, Courtney (NIH/NIDA) >>>>>>> [F] >>>>>>> wrote: >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> Hi, >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> I'm currently running analyses in Qdec and have identified >>>>>>>> areas that >>>>>>>> show significant effects in my contrasts. I have two questions >>>>>>>> related >>>>>>>> to interpreting these results. >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> 1. How can you interpret the effect in each "blob?" (i.e., >>>>>>>> which >>>>>>>> group has a greater cortical surface area, etc.). Right now, >>>>>>>> all I >>>>>>>> know is that there is an effect in certain areas. Can you >>>>>>>> export this >>>>>>>> data to get means and standard errors? >>>>>>>> >>>>>>>> 2. Is there a way to localize where these effects are? >>>>>>>> E.g., a >>>>>>>> tool that tells you this effect is located in __ area of the >>>>>>>> brain? >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> Thanks in advance >>>>>>>> >>>>>>>> Courtney >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> Courtney Gallen >>>>>>>> >>>>>>>> Post-baccalaureate IRTA >>>>>>>> >>>>>>>> Neuroimaging Research Branch >>>>>>>> >>>>>>>> National Institute on Drug Abuse (IRP) >>>>>>>> >>>>>>>> 251 Bayview Blvd >>>>>>>> >>>>>>>> Suite 200 >>>>>>>> >>>>>>>> Baltimore, MD 21224 >>>>>>>> >>>>>>>> Tel: (443) 740-2631tel:%28443%29%20740-2631 >>>>>>>> >>>>>>>> >>>>>>>> _______________________________________________ >>>>>>>> Freesurfer mailing list >>>>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> The information in this e-mail is intended only for the person >>>>>>> to whom it is >>>>>>> addressed. If you believe this e-mail was sent to you in error >>>>>>> and the e-mail >>>>>>> contains patient information, please contact the Partners >>>>>>> Compliance HelpLine at >>>>>>> http://www.partners.org/complianceline . If the e-mail was sent >>>>>>> to you in error >>>>>>> but does not contain patient information, please contact the >>>>>>> sender and properly >>>>>>> dispose of the e-mail. >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>> _______________________________________________ >>>>>> Freesurfer mailing list >>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>> >>>>>> _______________________________________________ >>>>>> Freesurfer mailing list >>>>>> Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu >>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Jesse C. Bledsoe, M.A. Ph.D. Candidate, Clinical Psychology Consortium for Neurodevelopmental Study Department of Psychology 110E Psychology Building Michigan State University East Lansing, MI 48824 ______________________________________
Note: Information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
freesurfer@nmr.mgh.harvard.edu
-
Douglas N Greve -
Gallen, Courtney (NIH/NIDA) [F] -
Michael Harms