search

FCP-INDI / fcp-indi.github.io

Github Pages Repo for FCP-INDI
https://fcp-indi.github.io
2 stars 8 forks source link

📝 Be more clear about the 36-parameter option and how to configure it #304

Open shnizzedy opened 1 year ago

shnizzedy commented 1 year ago

Related problem

can we please update our user guide to be more clear about the 36-parameter option and how to configure it? It's a very popular selection and we don't specifically highlight it in the guide.

📧

Proposed feature

Add a 36-parameter section to https://fcp-indi.github.io/docs/latest/user/nuisance?

Acceptance criteria

Alternatives

Add a 36-parameter tutorial to https://github.com/FCP-INDI/C-PAC_tutorials

Additional context

Currently the only place I see it mentioned is https://fcp-indi.github.io/docs/latest/user/release_notes/v1.4.1

sgiavasis commented 1 year ago

From the original paper* for more background information:

  • 3 parameters: No motion regression; only mean global timeseries, white matter (WM), and cerebrospinal fluid (CSF) timeseries were included (Fox et al., 2005). WM and CSF were defined on a subject-specific basis through segmentation of the T1-weighted image using FAST (Zhang et al., 2001). This model also forms the base model for comparisons with regression of voxelwise displacement parameters (below).

  • 9 parameters: standard 6 motion parameters (x, y, z translations and rotations) + WM/CSF/global timecourses.

  • 18 parameters: includes regressors from 9 parameter model, plus temporal derivative of each parameter across the timeseries (calculated using backward difference). The inclusion of the temporal derivative effectively accounts for a one-frame delay in the effect of motion on the BOLD signal. Such an 18-parameter model has been widely applied (Power et al., 2011; Van Dijk et al., 2011).

  • 36 parameters: additionally includes the quadratic term for all parameters in the 18-parameter model. Inclusion of the quadratic term effectively removes the sign of the motion parameter and also models nonlinearities in the effect of motion on the BOLD signal. This model is similar to the Volterra expansion proposed by Friston et al. (1996).

Satterthwaite TD, Elliott MA, Gerraty RT, Ruparel K, Loughead J, Calkins ME, Eickhoff SB, Hakonarson H, Gur RC, Gur RE, Wolf DH. An improved framework for confound regression and filtering for control of motion artifact in the preprocessing of resting-state functional connectivity data. Neuroimage. 2013 Jan 1;64:240-56. doi: 10.1016/j.neuroimage.2012.08.052. Epub 2012 Aug 25. PMID: 22926292; PMCID: PMC3811142.

shnizzedy commented 1 year ago

Related: https://github.com/FCP-INDI/fcp-indi.github.io/issues/138