History of AFNI updates  

|
April 10, 2014 01:01PM
Hi Gang,

Well, the results don't make sense in that there are no findings/clusters using 3dMEMA, whereas there were 18 clusters obtained with a different method (see below). For 3dMEMA, there were a few voxels, mostly outside the brain, and they don't seem to cluster.

I should mention that I am trying to use 3dMEMA to carry out a test of linear increases across three conditions using a conjunction analysis, so my final judgment about whether 3dMEMA is working correctly or not may not be right.

Can you verify that I am carrying out the conjunction analysis correctly?
First, I created two GLTs in 3dDeconvolve that coded for differences between conditions 1 and 2 and also conditions 2 and 3. Second, I tested each of these difference scores using the paired t-test example in 3dMEMA. Specifically, I put the 1 vs 2 difference score beta and t score into 3dMEMA and thresholded at p = 0.1. Then I did the same thing for the contrast 2 vs 3. Third, I found voxels that had positive t scores and that were common to both thresholded t-maps (conjunction analysis) using 3dcalc [step(thresholded tmap for conditions 1 vs 2) + step(thresholded tmap for conditions 2 vs 3)].

I am comparing the results of this 3dMEMA conjunction analysis to a test of linear increases across the three conditions using 3dRegAna. For this analysis, I code each condition with a number, i.e., cond 1 = 1, cond 2 = 2, cond 3 = 3, and I also code each subject with his/her own vector to make it a repeated measures analysis. The results of this analysis yield 18 clusters in the brain.

So, I am thinking that maybe I did not use 3dMEMA correctly or maybe I went wrong somewhere in my conjunction analysis.

One of my 3dMEMA commands is below, in case that might help.

Best,
Christine

3dMEMA -prefix SubjectiveFamiliarityHit6vs5 \
-jobs 16 \
-max_zeros 0.25 \
-missing_data 0 \
-set 6vs5 \
SS01 SS01/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS01/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS02 SS02/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS02/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS04 SS04/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS04/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS06 SS06/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS06/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS07 SS07/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS07/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS10 SS10/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS10/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS11 SS11/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS11/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS12 SS12/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS12/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS13 SS13/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS13/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS14 SS14/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS14/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS15 SS15/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS15/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS17 SS17/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS17/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]' \
SS18 SS18/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[3]' \
SS18/'decon_SubjectiveFamiliarityHit456_twottests_smartblur4_masked_ANTS+tlrc[4]'
Subject Author Posted

error message using 3dMEMA

Christine Smith April 09, 2014 07:32PM

Re: error message using 3dMEMA

gang April 10, 2014 11:05AM

Re: error message using 3dMEMA

Christine Smith April 10, 2014 01:01PM

Re: error message using 3dMEMA

Christine Smith April 10, 2014 01:46PM

Re: error message using 3dMEMA

gang April 10, 2014 01:54PM

Re: error message using 3dMEMA

Christine Smith April 10, 2014 04:04PM

Re: error message using 3dMEMA

gang April 10, 2014 04:36PM

Re: error message using 3dMEMA

Christine Smith April 11, 2014 12:37PM

Re: error message using 3dMEMA

gang April 11, 2014 03:48PM

Re: error message using 3dMEMA

Christine Smith April 14, 2014 12:35PM

Re: error message using 3dMEMA

gang April 14, 2014 04:12PM

Re: error message using 3dMEMA

Christine Smith April 14, 2014 04:57PM

Re: error message using 3dMEMA

Christine Smith April 14, 2014 05:16PM

Re: error message using 3dMEMA

gang April 14, 2014 05:31PM

Re: error message using 3dMEMA

Isaac Schwabacher April 10, 2014 02:02PM

Re: error message using 3dMEMA

Christine Smith April 10, 2014 02:38PM

Re: error message using 3dMEMA

Isaac Schwabacher April 10, 2014 03:19PM