You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
MRtrix MPPCA method (dwidenoise) now supports complex input data, which could help correcting for non-Gaussian noise biases present in magnitude-reconstructed MRI images.
Implementation considerations:
The NORDIC method could be adapted for other MRI modalities such as fMRI or ASL.
Implementing NORDIC similarly to the FIX tool used in micapipe fMRI (similar Matlab requirements?).
Test:
Code
# Input and output file paths
img_magn_in='/home/pabaua/dev_mni/data/test_nordic/sub-01_ses-2mm_dir-AP_run-01_part-mag_dwi.nii.gz'
img_phase_in='/home/pabaua/dev_mni/data/test_nordic/sub-01_ses-2mm_dir-AP_run-01_part-phase_dwi.nii.gz'
img_out='sub-01_ses-2mm_dir-AP_run-01_dwi_nordic' # Do not put extension .nii
nordic_matlab_dir='/home/pabaua/dev_mni/NORDIC_Raw/' # Directory containing NIFTI_NORDIC.m
# Arguments for the MATLAB function
ARG_temporal_phase=3
ARG_phase_filter_width=3
ARG_DIROUT='/home/pabaua/dev_mni/results/nordic/'
# Run MATLAB command with the specified arguments
matlab -nodisplay -nojvm -nosplash -nodesktop -r " \
try; \
addpath('${nordic_matlab_dir}'); \
ARG.temporal_phase = ${ARG_temporal_phase}; \
ARG.phase_filter_width = ${ARG_phase_filter_width}; \
ARG.DIROUT = '${ARG_DIROUT}'; \
NIFTI_NORDIC('${img_magn_in}', '${img_phase_in}', '${img_out}', ARG); \
end; \
quit;" \
>> ${ARG_DIROUT}/log_NORDIC_$(date '+%Y-%m-%d').txt
# Run dwidenoise on dMRI magnitude image
dwidenoise ${img_magn_in} ${ARG_DIROUT}/sub-01_ses-2mm_dir-AP_run-01_dwi_mppca.nii.gz
The text was updated successfully, but these errors were encountered:
PaulBautin
changed the title
implement NORDIC denoising in MicaPipe
NORDIC denoising in MicaPipe
Aug 26, 2024
What is the current behaviour?
Component Analysis) method, implemented in MRtrix3's
dwidenoise
command.Reference: https://mrtrix.readthedocs.io/en/dev/reference/commands/dwidenoise.html
What is the motivation / use case for changing the behaviour?
NORDIC references:
Before running:
Input data:
Alternative denoising method:
Reference: https://docs.dipy.org/stable/examples_built/preprocessing/denoise_patch2self.html
dwidenoise
) now supports complex input data, which could help correcting for non-Gaussian noise biases present in magnitude-reconstructed MRI images.Implementation considerations:
Test:
Code
The text was updated successfully, but these errors were encountered: