Module: preproc_dwi
Subworkflow performing pre-processing of DWI image from from brain extraction to motion correction and resampling. It requires at least one input channels including a DWI image with corresponding b-value (bval) and b-vector (bvec) files (ch_dwi). The next 3 channels are optional and includes a reversed DWI image with correspond bval and bvec files (ch_rev_dwi), a reversed phase encoded b = 0 image (ch_rev_b0) and a channel for Topup configuration file.
- Required channel include the following main steps: Brain Extraction, Denoising, Eddy, N4 Bias Correction, Normalization, and Resampling.
- Required and optional channels include the following main steps: Brain Extraction, Denoising, Topup+Eddy, N4 Bias Correction, Normalization and Resampling. The resulting file from this subworkflows is a DWI that has been denoised, corrected for susceptibility, eddy currents and motion, normalized, cropped and resampled. The next step would be to extract diffusion profiles for the corrected DWI, using a module implementing either DTI (see RECONST_DTI), fODF (see RECONST_FODF) or MAP-MRI (NotImplementedYet). --------- Steps -------------------- Brain Extraction (bet, FSL). Extract brain mask from the b0 image and applied to the whole DWI. This brain extraction is required to remove the skull and prepare the DWI to the T1 Registration. Denoising (dwidenoise, MPPCA method, MRtrix3). Used to remove the noise induced by the MRI acquisition, enhance the signal to noise ratio and improve the image quality and following metrics. The denoising is performed in the original spatial resolution and uses the MP-PCA method. Topup (FSL - optional). Topup uses the b=0 and reversed phase encoded b=0 images to extract the deformation field and corrects the brain deformation induced by the magnetic field susceptibility artefacts. Eddy (FSL). Eddy corrects eddy-currents, motion artefacts and performs slice-wise outlier detection and correction. When Topup is run, the eddy command is performed using the topup output. N4 Bias Correction (N4BiasFieldCorrection, ANTs). The N4 Bias Correction normalizes the image intensities and reduces this intensity bias (the center of the brain is less intense than its outer boundary due to multi-channel head coils). N4 correction is performed on the b=0 and applied to the whole DWI. Normalize (dwinormalise, MRtrix3). The DWI is normalized to have a mean value in the WM of approximately 1000. This task permits analyzing datasets from different MRI scanners with the same acquisition scheme. Resample (DIPY). The DWI is resampled to 1 mm isotropic spatial resolution, which is usually the spatial resolution of the T1. This spatial resolution is modifiable in the configuration file. See Tractoflow for more details, https://www.sciencedirect.com/science/article/pii/S105381192030375X?via%3Dihub
Inputs
| Type | Description | Pattern | |
|---|---|---|---|
| ch_dwi | file | The input channel containing the DWI file, B-values and B-vectors in FSL format files Structure: [ val(meta), path(dwi), path(bval), path(bvec) ] | *.{nii,nii.gz |
| ch_rev_dwi | file | The input channel containing the reverse DWI file, reverse B-values and reverse B-vectors in FSL format files Structure: [ val(meta), path(rev_dwi), path(bval), path(bvec) ] | *.{nii,nii.gz |
| ch_b0 | file | The input channel containing the b0 file. This input is optional. Structure: [ val(meta), path(rev_b0) ] | *.{nii,nii.gz} |
| ch_rev_b0 | file | The input channel containing the reverse b0 file. This input is optional. Structure: [ val(meta), path(rev_b0) ] | *.{nii,nii.gz} |
| ch_config_topup | file | The input channel containing the config file for Topup. This input is optional. See https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/topup/TopupUsersGuide#Configuration_files Structure: [ path(config_file) ] | *.{cnf} |
Outputs
| Type | Description | Pattern | |
|---|---|---|---|
| dwi_resample | file | Channel containing DWI denoised, corrected for susceptibility and/or eddy currents and motion, normalized, cropped and resampled. Structure: [ val(meta), path(dwi) ] | *_resampled.{nii,nii.gz} |
| bval | file | Channel containing Eddy-corrected B-values in FSL format Structure: [ val(meta), path(bval) ] | *__bval_eddy |
| bvec | file | Channel containing Eddy-corrected B-vectors in FSL format Structure: [ val(meta), path(bvec) ] | *__dwi_eddy_corrected.bvec |
| b0 | file | Channel containing b0 corrected file. Structure: [ val(meta), path(b0) ] | *__b0_bet.nii.gz |
| b0_mask | file | Channel containing b0 corrected binary mask file. Structure: [ val(meta), path(b0_mask) ] | *__b0_bet_mask.nii.gz |
| dwi_bounding_box | file | Channel containing the bounding box defining the limits of the crop. Structure: [ val(meta), path(bounding_box) ] | *.{pkl} |
| dwi_topup_eddy | file | Channel containing DWI output after denoised, correction for susceptibility and/or eddy currents and motion (Topup-Eddy). Structure: [ val(meta), path(dwi) ] | *_corrected.{nii,nii.gz} |
| dwi_n4 | file | Channel containing DWI after denoised, corrected for susceptibility and/or eddy currents and motion, cropped and normalized (N4). Structure: [ val(meta), path(dwi) ] | *_normalized.{nii,nii.gz} |
| versions | file | File containing software versions Structure: [ path(versions.yml) ] | versions.yml |
Components
| image/cropvolume |
| betcrop/fslbetcrop |
| denoising/mppca |
| image/resample |
| preproc/n4 |
| preproc/normalize |
| utils/extractb0 |
| preproc/gibbs |
| topup_eddy |
Keywords
| Brain extraction |
| Crop |
| Denoising |
| Topup-Eddy correction |
| Normalization |
| Resampling |
Authors
@medde