Normalized or Native MRI T1; zeropadding

[Joyce Chen]

Hello,

I'm just starting to use ROAST and had two questions:

1. In the first instance when running Roast, should we use normalized or native space T1 structural images. There are two possibilities:

a) run roast in native space, then normalize the outputs to standard MNI template to be able to compare data

b) run roast in normalized MNI space from the onset. But it seems that we may lose some information?

Is there one approach that is preferable?

2. When I first tried to run roast, it gave the error below. I then added the option zeropadding 30 and it worked. Is there any principled way to choose the padding value i.e. 10, 30, 60 etc. The manual says: "So do a'zeroPadding'of 10 to start with, and if you're not happy with the results, just increase the amount of zero-padding." For example, if both 30 and 60 values work, does this mean either would be fine, or is there an optimal value that I should find.

Thank you!

Joyce

Error message:

measuring head size...

Error using >=

Arrays have incompatible sizes for this operation.

Error in fitCap2individual (line 53)

temp_right_up = find((c_c>=c_c(indxinion))&(r_c<(r_c(I))));

Error in electrodePlacement (line 91)

[electrode_coord_P,center_P]= fitCap2individual(scalp,scalp_surface,landmarks,P2,capInfo,indP,isBiosemi,isEGI);

Error in roast (line 816)

hdrInfo = electrodePlacement(subj,subjRasRSPD,T2,elecName,options,

[Jake E Toth]

Hello Joyce,

1) You should run ROAST in native space then transform the results to MNI space (option a).
If you transform each head to MNI space first, then run ROAST each individuals results will look more or less the same.

2) As far as I am aware if you have very large zero padding then it makes the simulation take longer.
But if you have very small / no zero padding then sometimes the electrodes won't fit and you will get the error you got before.
So the ideal amount of zero padding is the smallest number you can use where you don't get an error.

I hope that helps!

Kind regards,
Jake Toth
PhD student in brain-computer interfaces at the University of Sheffield
jaketoth.tech 

[Joyce Chen]

Hi Jake,

Thanks very much! 

I have a few other questions, but I'll start a new email so that the subject reflects the questions. 

Kind regards,

Joyce