Multiple Q's: Resizing template, 3dtest++, clusterize settings defaults

AFNI Message Board
1

Hello. I’m new to analyzing fMRI data, and have been learning how to use AFNI. I also work with mouse data.

Is there a way to resize my template so I can overlay the obtained statistical map from deconvolve and ttest++ on it? I normalize my images using SPM, but when I display both the template and normalized image, the crosshair overlaps well. I’m assuming I can’t overlay it since the image dimensions are different. What option of 3drefit should I use if that’s the correct function? Also, is it possible to draw ROIs on my template and resize it to fit the dimensions of my normalized images as well?

As for 3dttest++, are the values uncorrected? When I use p<0.001, I barely get any activation appearing, while 0.05 gives me a lot with n=8. Currently, I am trying to find out if my normalization process is wrong or if I am using the wrong settings. I simply input my data under setA, apply mask, and calculate. I get two warnings when I run it. 3dttest++ set A:: 234 vectors are constant AND Smallest FDR q = 0.349733 ==> few true single voxel detections…

Also, is there a way to change the default settings under the clusterize menu? I want to default it to option 1 with 20 voxels. I couldn’t find such description in the environment readme.

Thank you

2

Probably better to ask distinct questions separately to get answers, but I’ll try here for the three-fer.

  1. Template and ROI resolution. You can usually resample with 3dresample. 3dNwarpApply can apply nonlinear warps with nearest neighbor resampling. 3dfractionize can be used to resample with control of the fraction of the voxel filled by an ROI. See our ROI documentation for more information. The @animal_warper program may be useful too for alignment to a mouse brain template and for moving atlas regions from template space to the native space of the subject mouse.

https://afni.nimh.nih.gov/pub/dist/edu/class_lectures/2020-03-NIH/afni_roi/afni11_roi.01.pdf

  1. 3dttest++. Results are uncorrected unless you use the -ClustSim or -ETAC options. You can also compute FWE cluster-corrected alpha values with 3dClustSim and 3dXClustSim. See the help for each of these programs for more information.

  2. Clusterize defaults. At present, there is no option to change those defaults. There are options to control other aspects of Clusterize though. See README.environment for variables starting with AFNI_CLUST…