eriosta / neuro

Dictionary Learning and GUI-based Labeling

This script performs fMRI data processing using the DictLearning algorithm from the nilearn package. It allows you to specify the number of subjects and components as command-line arguments.

Getting Started

Labeling App

usage: app.py [-h] [-s] n_subjects n_components

positional arguments:
  n_subjects    number of subjects
  n_components  number of components

options:
  -h, --help    show this help message and exit
  -s, --save    save '.nii' files

python app.py <N_SUBJECTS> <N_COMPONENTS>

python app.py 30 200

Viewer App

usage: viewer.py [-h] [--smoothing_fwhm SMOOTHING_FWHM] [--memory MEMORY] [--memory_level MEMORY_LEVEL] [--random_state RANDOM_STATE]
                 [--standardize STANDARDIZE] [--n_jobs N_JOBS]
                 n_subjects n_components method

positional arguments:
  n_subjects            number of subjects
  n_components          number of components
  method                decomposition method: 'dict_learning' or 'ica'

options:
  -h, --help            show this help message and exit
  --smoothing_fwhm SMOOTHING_FWHM
                        FWHM of Gaussian smoothing kernel
  --memory MEMORY       Directory to cache data
  --memory_level MEMORY_LEVEL
                        Level of memory caching
  --random_state RANDOM_STATE
                        Random state for reproducibility
  --standardize STANDARDIZE
                        Standardization method
  --n_jobs N_JOBS       Number of jobs to run in parallel

For Dictionary Learning

python viewer.py 10 20 dict_learning --smoothing_fwhm 6.0 --memory nilearn_cache --memory_level 2 --random_state 0 --standardize zscore_sample --n_jobs -1

For ICA

python viewer.py 10 20 ica --smoothing_fwhm 6.0 --memory nilearn_cache --memory_level 2 --random_state 0 --standardize zscore_sample --n_jobs -1

Prerequisites

To set up the environment for running the script, you have two options: creating a conda environment or installing the required libraries using pip.

Step 1

Install FMRIB Software Library (FSL)

• Download FSL

Step 2

Option 1: Conda Environment

Install Miniconda or Anaconda

• Download Miniconda
• Download Anaconda

Create a conda environment

conda env create -f environment.yml

Option 2: Pip Installation

Install Python

• Download Python

Install the required libraries using pip

pip install local_requirements.txt

Usage

Run the script using the following command:

python app.py <N_SUBJECTS> <N_COMPONENTS>

Replace <N_SUBJECTS> with the desired number of subjects and <N_COMPONENTS> with the desired number of components. For example, to process 5 subjects with 10 components each, use:

python app.py 5 10

Disclaimer

This script is provided as-is and without any warranty. Use it at your own risk. Ensure you have sufficient disk space and computational resources to perform the processing for the specified number of subjects and components.