For statistical datasets, you will need to set some of the attributes carefully. It's a little confusing because the information that describes the attributes in README.attributes describes the variables using a 0-based index, as for C. Matlab, on the other hand, uses a 1-based index, so you have to keep that in mind as you set the attributes.

In your case, the line that sets the BRICK_STATAUX attribute for an AFNI dataset is setting the attribute to a correlation coefficient by the first '2' in the line, which is inconsistent with the SCENE_DATA attribute, where you are setting the dataset to be a t-statistic (SCENE_DATA(2) could be 3 for t-statistic or 11 for functional bucket). The following command sets the number of parameters and then the parameters

InfoDelOut.BRICK_STATAUX = [1 2 3 160 2 2];


SCENE_DATA(1) = 0 means the dataset is in orig view, not 3DIM_HEAD_FUNC type as you put in the comments.
SCENE_DATA(3) = 2 means the dataset is 3DIM_GEN_ANAT to match the TYPESTRING attribute. To set this to make it 3DIM_HEAD_FUNC, use a value of 1 here instead and set OptDelOut.TYPESTRING to match as in the example below for your dataset.


DF = ; % put the degrees of freedom here
InfoDelOut.TYPESTRING = "3DIM_HEAD_FUNC" ;

InfoDelOut.SCENE_DATA(1) = 0; % orig view
InfoDelOut.SCENE_DATA(2) = 3; % fitt: t-statistic
InfoDelOut.SCENE_DATA(3) = 2; % 3DIM_HEAD_FUNC

InfoDelOut.BRICK_STATAUX = [0 3 1 DF ];