Show all posts by user
Dear AFNI users-
We are very pleased to announce that the new AFNI Message Board framework is up! Please join us at:
https://discuss.afni.nimh.nih.gov
Existing user accounts have been migrated, so returning users can login by requesting a password reset. New users can create accounts, as well, through a standard account creation process. Please note that these setup emails might initially go to spam folders (esp. for NIH users!), so please check those locations in the beginning.
The current Message Board discussion threads have been migrated to the new framework. The current Message Board will remain visible, but read-only, for a little while.
Sincerely,
AFNI HQ
History of AFNI updates
Results 2191 - 2220 of 2632
Steve,
The error indicates that the program failed in creating the header information about the output. Could you provide a little more about the error message, for example, copying from the place where you see
statpar <- "3drefit"
to the end?
by
gang
-
AFNI Message Board
Thomas, I'm not so sure what would happen if you don't specify a variable in the -model option. You can try it out with and without the reference, and compare the results. Also, 3dRegAna is obsolete now, and 3dttest++/3dMEMA is a better program for such analysis.
by
gang
-
AFNI Message Board
Elissa,
That's quite an impressive feat to have managed to run 3dttest++ the first time for a first-time AFNI user!
Regarding the 3dttest++ script: you should put quotes around the square brackets, as in, for example, EWA_2runs_H_P_bucket+tlrc'[26]' . Otherwise the program could just take the first sub-brick (number 0) as input. Also a better choice in specifying the sub-bri
by
gang
-
AFNI Message Board
> what functionalities AFNI has to deal with connectivity measures (e.g., betweenness, strength, efficiency) between regions.
Not sure what you're exactly looking for, but you may take a look at the structural vector autoregression approach here:
by
gang
-
AFNI Message Board
Elissa,
If you want to analyze the data for each taste separately, use 3dttest++ or 3dMEMA. If you want to include the three tastes as a factor in the model, try 3dMVM.
Gang
by
gang
-
AFNI Message Board
Michael,
If you are looking for the average or sum effect of two conditions, then the first test would be of interest to you. Otherwise the second test is the way to go.
by
gang
-
AFNI Message Board
Hi Mingbo,
> If I have used Matlab to generated some parameter (say, of the tuning curve of each voxel),
> and some value (say, the percentage of variance my tuning curve can explain for each voxel)
> that I want to use as threshold to visualize in AFNI, how should I assign those header info to
> these two types of data when I write data to BRIK files?
If you want to get it
by
gang
-
AFNI Message Board
> unequal spacing of TENT functions would be invaluable for block designs since
> the onset period is of course much more important than the boring plateau phase....
> do you think this is on the todo list for afni?
You're the first person who has asked for such an option. If more people want this, we may consider it.
> I am trying your suggestion of splitting it up into
by
gang
-
AFNI Message Board
> You mean that doing such a comparison could be meaningless or at least is hard to interpret?
I was just concerned about the interpretation. But if you don't think that is an issue, it should be fine.
> To get a firmer grasp on this I played a bit with the effect of scaling when adding effects
> of different sizes to the data, where two sets of data points are generated, one
by
gang
-
AFNI Message Board
> So then I would be running a t-test on values following a t-distribution...
> which is about fine with large n because then the t-distribution is similar
> to a normal distribution, isn't it?
Yes, that should be fine in asymptotic sense.
> To be clear, each map is not scaled by *its* sos (or sqrt(sos)). Rather the sos
> is computed over the maps (i.e. over subject
by
gang
-
AFNI Message Board
> How critical is the choice of duration for TENT(zero)/SPLIN(zero) etc. models?
It's really hard to pinpoint an exact number because several factors play roles about the response shape and length, and it may vary across cognitive states, stimulus types, tasks, brain regions, subjects, and groups. Too small a number would render an inaccurate characterization of the response, while ov
by
gang
-
AFNI Message Board
> I am not sure whether after scaling by the square root of sos they follow a t-distribution... should they?
Presumably they should follow a t-distribution by a factor, something like sqrt(n(n-1)).
> is it advisable to subtract the mean (of all z-scores) before computing the sos?
No, because you want to maintain the effect size of each map.
> I don't have a large number
by
gang
-
AFNI Message Board
Hi Nick,
Your wording of within- vs. between-maps confused me quite a bit before. Thanks for the clarifications.
> A direct comparison with 3dttest++ with -setA, setB and -paired gives a lot
> of blobs for the within>between just because it has higher mean values.
The interpretation of such a direct comparison is very straightforward. However, I see two potential issues with yo
by
gang
-
AFNI Message Board
> -model 5 : 1 2 3 0 -bucket 0 reg
It seems that you left out the 4th column, and it should be like this:
-model 5 : 0 1 2 3 4 -bucket 0 reg
> Now, in my output, the F-stat and R^2 would indicate the degree to which my
> explanatory data fits the data?
Yes, that's right.
> Or, would I look at the Coef # est & tstat of my explanatory variable for this? I ask
by
gang
-
AFNI Message Board
> There was a lot of significant activation outside of the brain
This is irrelevant.
> the results were completely different than we expected from previous comparison tests
> of the groups we are examining. Here is the script used in the 3dRegAna
A couple of potential problems:
1) Do the two genders have roughly the same average value for each of the three quantitative varia
by
gang
-
AFNI Message Board
Thomas,
> if I'm reading my output correctly then one of my covariates (scanner - data collected on two
> scanners) may be contributing an effect on this cluster above and beyond my variable A?
Did you center variable A? If not, the results for those two categorical variables would be interpreted as the variable A being at 0, which may or may not be meaningful. Could you provi
by
gang
-
AFNI Message Board
> Looking at the between- and within- maps separately (computed using 3dttest++ with -setA),
> the distribution of t-statistics seem quite similar.
I'm lost at your description. You have the correlation coefficients (transferred z-scores for group analysis) from individual subjects, right? When you say the distribution of t-statistics, it's the distribution of t-statistics (fr
by
gang
-
AFNI Message Board
> is there an option for unequal spacing of the intervals, i.e. to have more beta
> weights for the interesting first 20s than for the post-stimulus response?
One approach I can think of is to artificially break each block into two trial types (as well as their timings): the first has a duration of 20s, and the other starts at the end of the 20s block. Then you can arrange different bas
by
gang
-
AFNI Message Board
> I would like to plot (horizontal lines with the low frequencies regressed out, only stimulus related signal + noise)
You can add options -cbucket and -x1D in 3dDeconvolve to obtain all the regression coefficients and the model (design) matrix. Then get the effects of no interest:
3dSynthesize -cbucket CbucketFileFrom3dDeconvolve+orig -matrix x1DFrom3dDeconvolve -select ... -prefix Ef
by
gang
-
AFNI Message Board
Hi Thomas,
Regarding your analysis with 3dRegAna, do the age groups and two sexes have roughly the same average value of the explanatory variable A? What effects are you testing?
> Secondly, I have a separate dataset consisting of three groups: controls, patient responders,
> and patient nonresponders. I would like to find areas that discriminate between the groups in
> that nonr
by
gang
-
AFNI Message Board
> The ICC maps we generated from the ANOVA framework are strikingly similar
> to the ICC maps from 3dICC_REML for subjects...not session.
Michael,
Thanks for reporting this! I just found out a bug in 3dICC_REML a couple of weeks ago that may explain what you are seeing. In addition to the bug fix, I also modified the interface so that fixed effects are allowed in the model. See here
by
gang
-
AFNI Message Board
Jason,
I'm not so sure why FMRI power analysis would be dramatically different from the situation in other fields, nor do I see any need to write a special power analysis tool for FMRI data analysis. As long as you have enough information (effect size, standard deviation, desired significance level, power or sample size at the region of interest in the brain), you can always use online to
by
gang
-
AFNI Message Board
> The first regressor from the bucket output for condition1/condition2 should be
> the unmodulated amplitudes, however, we see very different results when
> comparing all-subjects 3dttest++ output from this data to data not using AM2.
> Basically, we see all the expected task activation when not using AM2 and
> almost none when using AM2. Any idea why we are seeing reduced acti
by
gang
-
AFNI Message Board
> What is the difference between using "-adiff 1 2" and usinf "-acontr 1 –1" in 3dANOVA2?
> I do not obtain the same activations when I use -adiff and -acontr. The activations are
> bigger (higher intensity) in -acontr than in -adiff. What is the reason for this difference?
That seems strange. Are the t-values the same? Could you provide the output of the followi
by
gang
-
AFNI Message Board
> In the documentation it indicates that the iresp option estimates the HRF at each
> voxel at each time point, but I don't know what the estimate means - is it the
> peak amplitude? the auc of the HRF?
Neither. The output from -iresp is the estimated response curve based on the weighted linear combination of the basis functions (SIN in your case) with weights being the beta v
by
gang
-
AFNI Message Board
Just to confirm: Do you have file R_io.so under /abin/afni?
ls -l /abin/R_io.so
by
gang
-
AFNI Message Board
> R however is 2.15.3 because when I updated to 3.0 I could no longer get MVM to run at all so I went back....
What error message did you get with R 3.0?
> -qVars 'PTAR+DPOEAR' \
Change it to
-qVars 'PTAR,DPOEAR' \
and see if it works.
by
gang
-
AFNI Message Board
What is your OS and its version?
What is your AFNI version? It can be found at the second line from the output of
afni -verb
by
gang
-
AFNI Message Board
There are two intriguing questions here:
1) Can one choose a relatively high voxel-wise p-threshold value (e.g., 0.1) as long the cluster-level significance level alpha) is controlled at 0.05?
Theoretically it should be fine because it's the cluster-level significance that matters more. A voxel-level p-value of 0.1 can be considered marginally significant. And if you adopt one-tailed t
by
gang
-
AFNI Message Board
> We have one value for each subject at each level.
Then you would have to go with the following: there is no way to differentiate MSAB and MSE, and you would simply consider MSAB as the within-subject MSE as I originally suggested with option -fa.
>> If yes, you use option -fb or -fab. With -fab, you will get two
>> sub-bricks with the first being MSB and the second
>
by
gang
-
AFNI Message Board