NingAnMe / DeepPrep-Data-Analysis

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

DeepPrep-Data-Analysis

denoise

The code in denoise dir is for DeepPrep 23.1.0 result.

Software dependence:

  1. FreeSurfer >= 6.0.0
  2. Python >= 3.10

Install Python dependence:

$ pip install bids numpy nibabel

Usage:

usage: bold_denoise.py [-h] --bold_preprocess_dir BOLD_PREPROCESS_DIR --bold_preproc_file BOLD_PREPROC_FILE [--repetition_time REPETITION_TIME] [--freesurfer_home FREESURFER_HOME] [--subjects_dir SUBJECTS_DIR]
                       --bold_denoise_dir BOLD_DENOISE_DIR

DeepPrep: denise

options:
  -h, --help            show this help message and exit
  --bold_preprocess_dir BOLD_PREPROCESS_DIR
                        DeepPrep BOLD result dir
  --bold_preproc_file BOLD_PREPROC_FILE
                        DeepPrep preprocessed BOLD file path
  --repetition_time REPETITION_TIME
                        RepetitionTime of BOLD file (optional)
  --freesurfer_home FREESURFER_HOME
                        freesurfer home path (optional)
  --subjects_dir SUBJECTS_DIR
                        DeepPrep Recon dir is required if the space of BOLD is fsnative (optional)
  --bold_denoise_dir BOLD_DENOISE_DIR
                        denoised BOLD file path

run example:

$ cd denoise

# Surface space
$ python3 bold_denoise.py --bold_preprocess_dir /DeepPrep_2310_Result/BOLD \
    --bold_preproc_file /DeepPrep_2310_Result/BOLD/sub-MSC01/ses-func01/func/sub-MSC01_ses-func01_task-rest_hemi-L_space-fsnative_bold.func.gii \
    --bold_denoise_dir /mnt/ngshare2/anning/workspace/DeepPrep-Data-Analysis/tmp/bold_denised_dir

# Volume space
$ python3 bold_denoise.py \ 
    --bold_preprocess_dir /DeepPrep_2310_Result/BOLD \
    --bold_preproc_file /DeepPrep_2310_Result/BOLD/sub-MSC01/ses-func01/func/sub-MSC01_ses-func01_task-rest_space-MNI152NLin6Asym_res-02_desc-preproc_bold.nii.gz \
    --bold_denoise_dir /mnt/ngshare2/anning/workspace/DeepPrep-Data-Analysis/tmp/bold_denised_dir \
    --repetition_time 2.2

About

License:Apache License 2.0


Languages

Language:Fortran 69.2%Language:Python 30.0%Language:Makefile 0.7%