 |
 |
Hi Mark, Yes, finally I can find the information that I need. Thank you so much, Buyean Lee -----Original Message----- From: Mark Jenkinson <[log in to unmask]> To: [log in to unmask] Sent: Mon, Nov 16, 2009 2:12 pm Subject: Re: [FSL] FIRST: critical F value Hi, I'm not entirely sure what you want here. You can create an uncorrect p-value mesh with the surface_fdr tool. This gives you the p-value corresponding to the F-value at each vertex. If you look at the corresponding vertex number in the two files then you can get the F and p values for that vertex. The thresholded versions are also available so that you can see if that vertex survived the FDR p-value threshold. Running surface_fdr also outputs the p-value threshold. So is this enough information to do what you want? If not, please explain exactly what you would like. All the best, Mark On 16 Nov 2009, at 20:47, Buyean Lee wrote: > Hi, > > I noticed a practical reason why I want to display the F-statistic > maps. > > Since the default values for 'Scalar range' in the Mesh Options are > '2 to 5' and also FSL 3D view can modify 'Mesh Options' for only the > first image opened, I can't figure out how to display both left and > right *_pvals.vtk of a structure (e.g., caudate) in the 'Scalar > range' of '0 to 0.05'. > > Again, I will appreciate if there is a way to calculate P-values for > some representative F-values in the *.vtk image (as described in the > paper by Zarei et al., NeuroImage 2009, Combining shape and > connectivity ...). > > Thanks a lot in advance, > > Buyean Lee > > > > -----Original Message----- > From: Buyean Lee <[log in to unmask]> > To: [log in to unmask] > Sent: Wed, Nov 4, 2009 10:35 pm > Subject: Re: [FSL] FIRST: critical F value > > Hi Mark, > > I want to display F-statistic map and describe some p-values (and > corresponding F statistics) as like in the Figure 1 of the paper by > Zarei et al.(NeuroImage 2009, Combining shape and connectivity ...). > > Thanks a lot, > > Buyean > > > -----Original Message----- > From: Mark Jenkinson <[log in to unmask]> > To: [log in to unmask] > Sent: Wed, Nov 4, 2009 2:15 pm > Subject: Re: [FSL] FIRST: critical F value > > Hi, > > I'm a bit confused here - do you want F stats or p values? > If you run surface_fdr then you get the multiple-comparison-> thresholded > F statistics as well as uncorrected p-values. You can certainly show > the F statistics from this very easily. Is that what you want? > > All the best, > Mark > > On 4 Nov 2009, at 21:44, Buyean Lee wrote: > > > Hi Mark, > > > > I understand that by running 'surface_fdr', one can convert F > > statistics to p-values. > > But, in a recent paper (Zarei et al., NeuroImage 2009, Combining > > shape and connectivity ...), I saw a case describing p-values for > > the corresponding F statistics (in Figure 1 legend) and presenting F-> > statistic maps. > > > > I might be too picky, but I prefer presenting a F-statistic map > > because it seems easier to display high F-values (smaller p-values) > > in F-statistic maps. > > > > Is there any way to calculate p-values for the corresponding F > > statistics? > > > > Thank you, > > > > Buyean > > > > > > -----Original Message----- > > From: Mark Jenkinson <[log in to unmask]> > > To: [log in to unmask] > > Sent: Thu, Sep 3, 2009 2:42 pm > > Subject: Re: [FSL] FIRST: critical F value > > > > Hi, > > > > Just run surface_fdr - it will give you uncorrected p-values which > > you can > > easily display and threshold plus FDR-corrected F-stats which are > > thresholded. > > See http://www.fmrib.ox.ac.uk/fsl/first/index.html for more info. > > > > All the best, > > Mark > > > > On 3 Sep 2009, at 19:38, Buyean Lee wrote: > > > > > Dear FSL FIRST users, > > > > > > I am trying to display the results of vertex analysis with only > > > those regions with F values that are significant (for now > > > uncorrected for multiple comparison correction). I think I need to > > > put the critical F value in the Salar range of Mesch Options, but > I > > don't know how to find this critical F-value. > > > > > > Any suggestion? > > > > > > Thank you, > > > > > > Buyean > > > > > >