Slice Timing Correction

[BrightMG]

Add slice timing correction (3dTshift) option

(When should this occur? After RETROICOR and volume registration?)

[RayStick]

https://www.frontiersin.org/articles/10.3389/fnins.2019.00821/full

[RayStick]

The article I posted above explores when in your processing is best to do slice timing correction.

@kristinazvolanek at the moment, me & you haven't been doing this in our analyses. For your current data, being MB8 and a very fast TR of 0.555s, it will probably make a minimal change, and you could easily get away without doing it.

For my data, MB4 and TR 1.2 seconds, there may be more rationale to try and implement this? But I'm not fully sure how necessary it is.

[RayStick]

3dTshift has a lot of warnings/caveats with it.

[kristinazvolanek]

I think we should have a closer look at this article and maybe meet to discuss? Although at the beginning they say slice timing correction is often not used for short TRs, they report later that STC significantly improved the correlation between a voxel's timeseries and its regressor. From a quick skim it seems that the decision to correct for slice timing might also depend on how we implement smoothing. I know this was another point of discussion but can't remember exactly where we ended up on that decision.