Dear AFNI experts,

I would like to investigate the relationship between structural co-variance networks and individual differences in verbal working memory by conducting analyses similar to the ones found in Bernhardt et al. 2013.

In Bernhardt et al. 2013, the authors used the SurfStat toolbox (Worsley et al. 2009) to:

1) Map the structural co-variance networks by cortical thickness correlations

The authors fitted the model for thickness T at surface point i.

T_i = B₀ + B₁T_seed

The authors used the CT of two seeds (dorsal anterior insula and anterior midcingulate cortex)

2) Assess the relationship between co-variance strength and a behavioral score (i.e. empathy ratings)

The authors fitted

T_i= B₀ + B₁T_seed + B₂BehavioralScore + B₃(Behavioral Score X T_seed).

I would like to conduct a similar analysis in AFNI. I have already exported the CT map of each participant in SUMA.

Previously, I have used 3dTtest++ to investigate the relationship between CT and behavioral scores. I can see how I could use 3dTtest++ to investigate the relationship between CT across the whole-brain and the CT in a seed region.

Within 3dTtest++, I would include as covariates of no interest (age and overall CT) and as covariates of interest the CT of my seed region.

However, I am not sure how I could assess the relationship between co-variance strength and a behavioral score using AFNI (B₃(Behavioral Score X T_seed)).

Any help would be most welcomed.

Isa