FSL Component Regression Filter (fsl_regfilt)¶
Library: FSL | Docker Image: brainlife/fsl
Function¶
Removes nuisance components from 4D fMRI data by regressing out specified columns from a design or mixing matrix.
Modality: 4D fMRI NIfTI time series with associated ICA mixing matrix.
Typical Use: Removing ICA-identified noise components from fMRI data, often used with FIX or manual classification.
Key Parameters¶
-i (input 4D), -d (design/mixing matrix), -o (output), -f (component indices to remove)
Key Points¶
Typically used with MELODIC output. Removes ICA components classified as noise. Component indices are comma-separated or ranges.
Inputs¶
| Name | Type | Required | Label | Flag |
|---|---|---|---|---|
input |
File |
Yes | Input 4D data file | -i |
design |
File |
Yes | Design matrix (e.g., MELODIC mixing matrix) | -d |
output |
string |
Yes | Output filename | -o |
filter |
string |
Yes | Component indices to filter out (comma-separated, 1-indexed) | -f |
aggressive |
boolean |
No | Use aggressive (full variance) filtering | -a |
freq_filter |
boolean |
No | Frequency-based filtering | --freq |
freq_ic |
File |
No | Frequency IC file for frequency-based filtering | --freq_ic |
verbose |
boolean |
No | Verbose output | -v |
mask |
File |
No | Brain mask | -m |
Accepted Input Extensions¶
- input:
.nii,.nii.gz - design:
.txt,.mat
Outputs¶
| Name | Type | Glob Pattern |
|---|---|---|
filtered_data |
File |
$(inputs.output).nii.gz, $(inputs.output).nii, $(inputs.output) |
log |
File |
fsl_regfilt.log |
Output Extensions¶
- output:
.nii,.nii.gz
Docker Tags¶
Available versions: latest, 6.0.4-patched2, 6.0.4-patched, 6.0.4, 6.0.4-xenial, 5.0.11, 6.0.0, 6.0.1, 5.0.9
Categories¶
- Functional MRI > FSL > ICA/Denoising