Inverse Logit

[Jenni]

Dear All,

Thank you so much for making all these wonderful tools available!

We are trying to use the Inverse Logit model to model our data. I've used the sFIR_multisubject_example_script.m as a template to do this. I changed line 125 to
[params_obj, hrf_obj] = hrf_fit(d, TR, C, T, 'IL', 0);
in order to use the IL model instead of sFIR.

My worry is when I run it on one subject with a small mask just to test it, the results look a bit odd. When I plot any of the HRF_timeseries_condition.img images there are a lot of straight lines.

Here is a link to what the timeseries plot looks like:
https://www.dropbox.com/s/348bbxi5ka77nv0/IL_condition_1.jpg?dl=0

The reason I don't think this is right is because when I run the same script with sFIR as the basis set, the resulting HRF has much more sensible shape. It basically looks more like what you would expect the HRF to look like.

So, I wanted to check can I use the sFIR_multisubject_example_script.m to model my data with IL? If not, can you tell me if there is a way to modify the script to run IL?

Many thanks,
Jenni

[Stephan Geuter]

Hi Jenni,

The code should be working the the IL response function, too. The plot you’re sharing, are these the modeled time-courses for sFIR and IL models, respectively?

Best,
Stephan

> --
> You received this message because you are subscribed to the Google Groups "WagerlabTools" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to wagerlabtool...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[jenni0l...@gmail.com]

Hi Stephan,

Sorry! That wasn't a ver good or informative figure. It was just the IL model and the second plot was empty.

I tried it with different data in case that was it but the results are pretty much the same. Here is the sFIR and IL for same next to each other:

https://www.dropbox.com/s/nmdi3plj09dabbh/Condition_1_sFIR_IL.jpg?dl=0

I find it's very odd that in the IL model the HRF y axis goes from 0 to 90, while in the sFIR model it's -0.3 - 0.4. I must have missed something?

Many thanks,
Jenni