Yes, just specify the age^2 as a covariate (in addition to the the
linear trend). Add age^3 as well. Note that you should demean each, and
probably rescale so that the max of each is 1.
Dalwani, Manish wrote:
> Hi All,
>
> If you don't have a linear trend in the Cortical thickness vs age
> co
Hi All,
If you don't have a linear trend in the Cortical thickness vs age correlation,
is there a way to use a polynomial (quadratic, cubic) regression model to best
fit the cortical thickness vs age using the GLM mechanism of freesurfer?
Regards,
Manish Dalwani
Sr. PRA
Univ. of CO
