Metadata-Version: 2.4
Name: MEGnet-neuro
Version: 0.3.1
Summary: Package to calculate and classify ICAs using MEGNET deep learning architecture
Author-email: Jeff Stout <stoutjd@nih.gov>, Allison Nugent <nugenta@nih.gov>
License-Expression: LicenseRef-LICENSE
Classifier: Programming Language :: Python :: 3
Classifier: Operating System :: POSIX
Requires-Python: <3.12
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: mne
Requires-Dist: numpy
Requires-Dist: scipy
Requires-Dist: pandas
Requires-Dist: munch
Requires-Dist: nibabel
Requires-Dist: joblib
Requires-Dist: tensorflow==2.15
Requires-Dist: scikit-learn
Requires-Dist: huggingface_hub>=0.24.0
Provides-Extra: dev
Requires-Dist: stabilized-ica; extra == "dev"
Provides-Extra: testing
Requires-Dist: datalad; extra == "testing"
Requires-Dist: pytest; extra == "testing"
Requires-Dist: pygit2; extra == "testing"
Dynamic: license-file

# MEGNET
[![megnet-tests](https://github.com/nih-megcore/MegNet/actions/workflows/megnet-actions.yml/badge.svg)](https://github.com/nih-megcore/MegNet/actions/workflows/megnet-actions.yml)

This repository is a fork of the code listed below in the original code reference.  This repository adds an automated processing wrapper and python package installation around the original codebase.  The current codebase utilizes mne python to preprocess the data, generate the infomax ICA components (n=20), circular topography maps, and timeseries outputs.  The architecture of neural net has been preserved, however, the weights have been reset to uniform distribution and retrained using repository data from MEGIN, CTF, 4D, and KIT systems.

## Install
```
conda create -n megnet 'mne>=1.6' 'python<3.12'
conda activate megnet
pip install git+https://github.com/nih-megcore/MegNet.git
```

## Post install initalization (downloading huggingface model weights)
```
megnet_init
```


## Original Code Repository
https://github.com/DeepLearningForPrecisionHealthLab/MegNET_2020 <br>
Manuscript available: https://pubmed.ncbi.nlm.nih.gov/34274419/ <br>
DOI: https://doi.org/10.1016/j.neuroimage.2021.118402 <br>
PMID: 34274419 <br>

