There is nothing to fit nonlinear parameters. You can do some simpling things. Eg, if you believe that thickness is changing exponentially with time, th = exp(-a/t), then you can take the log of the thickness so that log(th) = -a/t, then fit for a.

doug

mmoayedi wrote:

Hi Everyone,

I was wondering if there's a method to examine non-linear correlations in FreeSurfer. Specifically, I'd like to use a behavioural correlate and see if there are any correlations in the cortex. The linear correlation wasn't as fruitful as I'd hoped, but there is evidence that a non-linear relationship may exist. Is there any way of testing this?

Thanks.

Regards,

Massieh Moayedi PhD Candidate Dr. Karen Davis' Lab Institute of Medical Science University of Toronto

MP14-303 Brain, Imaging and Behaviour Toronto Western Research Institute University Health Network 399 Bathurst St. Toronto, ON M5T 2S8

massih.moayedi@utoronto.ca mmoayedi@uhnres.utoronto.ca

t: 416-603-5502 f: 416-603-5745

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer