 |
 |
Hi Michael, please find the .png and .fsf files attached. The content of the gfeat/cope2 (and all subsequent copes) is a log folder, a stats folder, design files, report files, example_func, mean_func, and a mask. In the stats folder, there is only a log and a DoF file. Given that you picked up my initial question, I assume that you generally agree with the procedure that I suggested (to use zstats maps from the second level feat as randomise inputs for a 2x2 ANOVA). I am asking because the contrast set up in FSL is a little confusing for me. I am a BrainVoyager turned FSL user and am used to set up all contrasts on the group level. So the fact that I am setting up contrasts on the level of individual runs is strange to me. I know how to use randomise to set up the ANOVA that I want to do and therefore planned my analysis this way. But if you think that I should do this differently (i.e., set up the ANOVA as a higher level feat) please let me know. Maren On Mon, May 20, 2013 at 9:48 AM, Harms, Michael <[log in to unmask]> wrote: > Can you post your .png and design.fsf files? Also, what exactly are the > contents of the .gfeat/cope2.feat directory? > > -- > Michael Harms, Ph.D. > ----------------------------------------------------------- > Conte Center for the Neuroscience of Mental Disorders > Washington University School of Medicine > Department of Psychiatry, Box 8134 > 660 South Euclid Ave. Tel: 314-747-6173 > St. Louis, MO 63110 Email: [log in to unmask] > > From: Maren Strenziok <[log in to unmask]> > Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> > Date: Saturday, May 18, 2013 1:03 PM > > To: FSL - FMRIB's Software Library <[log in to unmask]> > Subject: Re: [FSL] Feat second level zstats maps > > Hi, > > I tried both, fixed and mixed effects (EVs set as you wrote), but I stil > lonly get one zstat map (for cope1). > > Maren > > > On Sat, May 18, 2013 at 8:46 AM, Harms, Michael <[log in to unmask]> wrote: > >> >> Well, then I'm confused, because what you wrote that you did in your >> 2nd level analysis is not what I described to do. (i.e., use Fixed, not >> Mixed effects; and what is an EV set to 0?) >> >> Did you use the "Inputs are lower level FEAT directories" setting on >> the Data tab? >> >> cheers, >> -MH >> >> -- >> Michael Harms, Ph.D. >> ----------------------------------------------------------- >> Conte Center for the Neuroscience of Mental Disorders >> Washington University School of Medicine >> Department of Psychiatry, Box 8134 >> 660 South Euclid Ave. Tel: 314-747-6173 >> St. Louis, MO 63110 Email: [log in to unmask] >> >> From: Maren Strenziok <[log in to unmask]> >> Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> >> Date: Friday, May 17, 2013 9:35 PM >> >> To: FSL - FMRIB's Software Library <[log in to unmask]> >> Subject: Re: [FSL] Feat second level zstats maps >> >> That's what I did. It didn't give me a zstat map for each COPE, just >> for the first one. The question that I posted was how to get a zstat map >> for each on my COPEs. >> >> Maren >> >> >> On Fri, May 17, 2013 at 6:32 PM, Harms, Michael <[log in to unmask]> wrote: >> >>> >>> If you 2nd level analysis is just to effectively average the 1st level >>> copes from the 5 runs in a given subject, then you should do a Fixed >>> Effects Lev2 analysis using a single EV that is a column of 1's, and then >>> you'll get a .gfeat directory for each subject with copeX.feat directories, >>> where X is all your 1st level contrasts. >>> >>> cheers, >>> -MH >>> >>> -- >>> Michael Harms, Ph.D. >>> ----------------------------------------------------------- >>> Conte Center for the Neuroscience of Mental Disorders >>> Washington University School of Medicine >>> Department of Psychiatry, Box 8134 >>> 660 South Euclid Ave. Tel: 314-747-6173 >>> St. Louis, MO 63110 Email: [log in to unmask] >>> >>> From: Maren Strenziok <[log in to unmask]> >>> Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> >>> Date: Friday, May 17, 2013 2:27 PM >>> To: FSL - FMRIB's Software Library <[log in to unmask]> >>> Subject: [FSL] Feat second level zstats maps >>> >>> Hi, >>> >>> I am planning an fMRI data analysis using FEAT. I want to look at a >>> pre-post-intervention change between two groups that were trained under two >>> different conditions. At each session, 5 runs of data were acquired. >>> Roughly, what I am planning is to run a first-level and a second-level feat >>> analysis and take the z-stats maps from the second level analysis to do a >>> group x time analysis using randomise. I got stuck at the second level. >>> Also, if there are objections to do this this way, I would appreciate any >>> suggestions. >>> 1. In the first-level feat analysis of individual runs, I set up >>> standard preprocessing parameters and registered the functional data to MNI >>> 1mm space via a high-resolution mprage. In the stats tab, I set up 1 EV for >>> each predictor. In the contrasts, I set up 1 contrast for each EV and a few >>> more additional contrasts to look at differences between conditions in my >>> experiment. >>> 2. In the second-level analysis, I used the lower-level feat directories >>> as inputs. Under stats, I selected Mixed effects, FLAME1. In the Full model >>> set up, I set up one EV (set to 0). This gave me only a zstat map for the >>> COPE of my first EV, whereas I got only log and DoF files for the other 16 >>> EVs. I tried to set up 17 EVs, but that gave me an error message. How do I >>> get a zstat map for each of the COPEs or should I not do this this way at >>> all? >>> >>> Maren >>> >>> >>> ------------------------------ >>> >>> The materials in this message are private and may contain Protected >>> Healthcare Information or other information of a sensitive nature. If you >>> are not the intended recipient, be advised that any unauthorized use, >>> disclosure, copying or the taking of any action in reliance on the contents >>> of this information is strictly prohibited. If you have received this email >>> in error, please immediately notify the sender via telephone or return mail. >>> >> >> >> ------------------------------ >> >> The materials in this message are private and may contain Protected >> Healthcare Information or other information of a sensitive nature. If you >> are not the intended recipient, be advised that any unauthorized use, >> disclosure, copying or the taking of any action in reliance on the contents >> of this information is strictly prohibited. If you have received this email >> in error, please immediately notify the sender via telephone or return mail. >> > > > ------------------------------ > > The materials in this message are private and may contain Protected > Healthcare Information or other information of a sensitive nature. If you > are not the intended recipient, be advised that any unauthorized use, > disclosure, copying or the taking of any action in reliance on the contents > of this information is strictly prohibited. If you have received this email > in error, please immediately notify the sender via telephone or return mail. > On Mon, May 20, 2013 at 9:48 AM, Harms, Michael <[log in to unmask]> wrote: > Can you post your .png and design.fsf files? Also, what exactly are the > contents of the .gfeat/cope2.feat directory? > > -- > Michael Harms, Ph.D. > ----------------------------------------------------------- > Conte Center for the Neuroscience of Mental Disorders > Washington University School of Medicine > Department of Psychiatry, Box 8134 > 660 South Euclid Ave. Tel: 314-747-6173 > St. Louis, MO 63110 Email: [log in to unmask] > > From: Maren Strenziok <[log in to unmask]> > Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> > Date: Saturday, May 18, 2013 1:03 PM > > To: FSL - FMRIB's Software Library <[log in to unmask]> > Subject: Re: [FSL] Feat second level zstats maps > > Hi, > > I tried both, fixed and mixed effects (EVs set as you wrote), but I stil > lonly get one zstat map (for cope1). > > Maren > > > On Sat, May 18, 2013 at 8:46 AM, Harms, Michael <[log in to unmask]> wrote: > >> >> Well, then I'm confused, because what you wrote that you did in your >> 2nd level analysis is not what I described to do. (i.e., use Fixed, not >> Mixed effects; and what is an EV set to 0?) >> >> Did you use the "Inputs are lower level FEAT directories" setting on >> the Data tab? >> >> cheers, >> -MH >> >> -- >> Michael Harms, Ph.D. >> ----------------------------------------------------------- >> Conte Center for the Neuroscience of Mental Disorders >> Washington University School of Medicine >> Department of Psychiatry, Box 8134 >> 660 South Euclid Ave. Tel: 314-747-6173 >> St. Louis, MO 63110 Email: [log in to unmask] >> >> From: Maren Strenziok <[log in to unmask]> >> Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> >> Date: Friday, May 17, 2013 9:35 PM >> >> To: FSL - FMRIB's Software Library <[log in to unmask]> >> Subject: Re: [FSL] Feat second level zstats maps >> >> That's what I did. It didn't give me a zstat map for each COPE, just >> for the first one. The question that I posted was how to get a zstat map >> for each on my COPEs. >> >> Maren >> >> >> On Fri, May 17, 2013 at 6:32 PM, Harms, Michael <[log in to unmask]> wrote: >> >>> >>> If you 2nd level analysis is just to effectively average the 1st level >>> copes from the 5 runs in a given subject, then you should do a Fixed >>> Effects Lev2 analysis using a single EV that is a column of 1's, and then >>> you'll get a .gfeat directory for each subject with copeX.feat directories, >>> where X is all your 1st level contrasts. >>> >>> cheers, >>> -MH >>> >>> -- >>> Michael Harms, Ph.D. >>> ----------------------------------------------------------- >>> Conte Center for the Neuroscience of Mental Disorders >>> Washington University School of Medicine >>> Department of Psychiatry, Box 8134 >>> 660 South Euclid Ave. Tel: 314-747-6173 >>> St. Louis, MO 63110 Email: [log in to unmask] >>> >>> From: Maren Strenziok <[log in to unmask]> >>> Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> >>> Date: Friday, May 17, 2013 2:27 PM >>> To: FSL - FMRIB's Software Library <[log in to unmask]> >>> Subject: [FSL] Feat second level zstats maps >>> >>> Hi, >>> >>> I am planning an fMRI data analysis using FEAT. I want to look at a >>> pre-post-intervention change between two groups that were trained under two >>> different conditions. At each session, 5 runs of data were acquired. >>> Roughly, what I am planning is to run a first-level and a second-level feat >>> analysis and take the z-stats maps from the second level analysis to do a >>> group x time analysis using randomise. I got stuck at the second level. >>> Also, if there are objections to do this this way, I would appreciate any >>> suggestions. >>> 1. In the first-level feat analysis of individual runs, I set up >>> standard preprocessing parameters and registered the functional data to MNI >>> 1mm space via a high-resolution mprage. In the stats tab, I set up 1 EV for >>> each predictor. In the contrasts, I set up 1 contrast for each EV and a few >>> more additional contrasts to look at differences between conditions in my >>> experiment. >>> 2. In the second-level analysis, I used the lower-level feat directories >>> as inputs. Under stats, I selected Mixed effects, FLAME1. In the Full model >>> set up, I set up one EV (set to 0). This gave me only a zstat map for the >>> COPE of my first EV, whereas I got only log and DoF files for the other 16 >>> EVs. I tried to set up 17 EVs, but that gave me an error message. How do I >>> get a zstat map for each of the COPEs or should I not do this this way at >>> all? >>> >>> Maren >>> >>> >>> ------------------------------ >>> >>> The materials in this message are private and may contain Protected >>> Healthcare Information or other information of a sensitive nature. If you >>> are not the intended recipient, be advised that any unauthorized use, >>> disclosure, copying or the taking of any action in reliance on the contents >>> of this information is strictly prohibited. If you have received this email >>> in error, please immediately notify the sender via telephone or return mail. >>> >> >> >> ------------------------------ >> >> The materials in this message are private and may contain Protected >> Healthcare Information or other information of a sensitive nature. If you >> are not the intended recipient, be advised that any unauthorized use, >> disclosure, copying or the taking of any action in reliance on the contents >> of this information is strictly prohibited. If you have received this email >> in error, please immediately notify the sender via telephone or return mail. >> > > > ------------------------------ > > The materials in this message are private and may contain Protected > Healthcare Information or other information of a sensitive nature. If you > are not the intended recipient, be advised that any unauthorized use, > disclosure, copying or the taking of any action in reliance on the contents > of this information is strictly prohibited. If you have received this email > in error, please immediately notify the sender via telephone or return mail. > On Mon, May 20, 2013 at 9:48 AM, Harms, Michael <[log in to unmask]> wrote: > Can you post your .png and design.fsf files? Also, what exactly are the > contents of the .gfeat/cope2.feat directory? > > -- > Michael Harms, Ph.D. > ----------------------------------------------------------- > Conte Center for the Neuroscience of Mental Disorders > Washington University School of Medicine > Department of Psychiatry, Box 8134 > 660 South Euclid Ave. Tel: 314-747-6173 > St. Louis, MO 63110 Email: [log in to unmask] > > From: Maren Strenziok <[log in to unmask]> > Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> > Date: Saturday, May 18, 2013 1:03 PM > > To: FSL - FMRIB's Software Library <[log in to unmask]> > Subject: Re: [FSL] Feat second level zstats maps > > Hi, > > I tried both, fixed and mixed effects (EVs set as you wrote), but I stil > lonly get one zstat map (for cope1). > > Maren > > > On Sat, May 18, 2013 at 8:46 AM, Harms, Michael <[log in to unmask]> wrote: > >> >> Well, then I'm confused, because what you wrote that you did in your >> 2nd level analysis is not what I described to do. (i.e., use Fixed, not >> Mixed effects; and what is an EV set to 0?) >> >> Did you use the "Inputs are lower level FEAT directories" setting on >> the Data tab? >> >> cheers, >> -MH >> >> -- >> Michael Harms, Ph.D. >> ----------------------------------------------------------- >> Conte Center for the Neuroscience of Mental Disorders >> Washington University School of Medicine >> Department of Psychiatry, Box 8134 >> 660 South Euclid Ave. Tel: 314-747-6173 >> St. Louis, MO 63110 Email: [log in to unmask] >> >> From: Maren Strenziok <[log in to unmask]> >> Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> >> Date: Friday, May 17, 2013 9:35 PM >> >> To: FSL - FMRIB's Software Library <[log in to unmask]> >> Subject: Re: [FSL] Feat second level zstats maps >> >> That's what I did. It didn't give me a zstat map for each COPE, just >> for the first one. The question that I posted was how to get a zstat map >> for each on my COPEs. >> >> Maren >> >> >> On Fri, May 17, 2013 at 6:32 PM, Harms, Michael <[log in to unmask]> wrote: >> >>> >>> If you 2nd level analysis is just to effectively average the 1st level >>> copes from the 5 runs in a given subject, then you should do a Fixed >>> Effects Lev2 analysis using a single EV that is a column of 1's, and then >>> you'll get a .gfeat directory for each subject with copeX.feat directories, >>> where X is all your 1st level contrasts. >>> >>> cheers, >>> -MH >>> >>> -- >>> Michael Harms, Ph.D. >>> ----------------------------------------------------------- >>> Conte Center for the Neuroscience of Mental Disorders >>> Washington University School of Medicine >>> Department of Psychiatry, Box 8134 >>> 660 South Euclid Ave. Tel: 314-747-6173 >>> St. Louis, MO 63110 Email: [log in to unmask] >>> >>> From: Maren Strenziok <[log in to unmask]> >>> Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> >>> Date: Friday, May 17, 2013 2:27 PM >>> To: FSL - FMRIB's Software Library <[log in to unmask]> >>> Subject: [FSL] Feat second level zstats maps >>> >>> Hi, >>> >>> I am planning an fMRI data analysis using FEAT. I want to look at a >>> pre-post-intervention change between two groups that were trained under two >>> different conditions. At each session, 5 runs of data were acquired. >>> Roughly, what I am planning is to run a first-level and a second-level feat >>> analysis and take the z-stats maps from the second level analysis to do a >>> group x time analysis using randomise. I got stuck at the second level. >>> Also, if there are objections to do this this way, I would appreciate any >>> suggestions. >>> 1. In the first-level feat analysis of individual runs, I set up >>> standard preprocessing parameters and registered the functional data to MNI >>> 1mm space via a high-resolution mprage. In the stats tab, I set up 1 EV for >>> each predictor. In the contrasts, I set up 1 contrast for each EV and a few >>> more additional contrasts to look at differences between conditions in my >>> experiment. >>> 2. In the second-level analysis, I used the lower-level feat directories >>> as inputs. Under stats, I selected Mixed effects, FLAME1. In the Full model >>> set up, I set up one EV (set to 0). This gave me only a zstat map for the >>> COPE of my first EV, whereas I got only log and DoF files for the other 16 >>> EVs. I tried to set up 17 EVs, but that gave me an error message. How do I >>> get a zstat map for each of the COPEs or should I not do this this way at >>> all? >>> >>> Maren >>> >>> >>> ------------------------------ >>> >>> The materials in this message are private and may contain Protected >>> Healthcare Information or other information of a sensitive nature. If you >>> are not the intended recipient, be advised that any unauthorized use, >>> disclosure, copying or the taking of any action in reliance on the contents >>> of this information is strictly prohibited. If you have received this email >>> in error, please immediately notify the sender via telephone or return mail. >>> >> >> >> ------------------------------ >> >> The materials in this message are private and may contain Protected >> Healthcare Information or other information of a sensitive nature. If you >> are not the intended recipient, be advised that any unauthorized use, >> disclosure, copying or the taking of any action in reliance on the contents >> of this information is strictly prohibited. If you have received this email >> in error, please immediately notify the sender via telephone or return mail. >> > > > ------------------------------ > > The materials in this message are private and may contain Protected > Healthcare Information or other information of a sensitive nature. If you > are not the intended recipient, be advised that any unauthorized use, > disclosure, copying or the taking of any action in reliance on the contents > of this information is strictly prohibited. If you have received this email > in error, please immediately notify the sender via telephone or return mail. >