ENIGMA MEG Pipeline FAQ
Under construction -- If there is a topic you would like added, reach out to Anna Namyst. :)
The programs in this package perform the full processing pipeline for the ENIGMA BIDS working group. This suite requires that your data be organized in BIDS format
Preparation
- Why BIDS?
- Brain Imaging Data Structure (BIDS) provides a standard for organizing neuroimaging data.
- My data isn't in BIDS -- how can I get it into that format?
- If you need a tool for that you can use enigma_anonymization_lite. The core tool is process_meg.py, which performs all processing steps for the anatomical MRI and the associated MEG. You can either process a single subject, or you can loop over all subjects in batch mode.
- How do I batch process my data?
- In order to do batch processing, you must first run parse_bids.py to produce a .csv file manifest of all available MEG scans (or use some other method to generate the .csv file).
- Does the pipeline include artifact correction?
- Yes, there are two methods of artifact correction supported. The first is ica with manual identification of ica components. This is likely the most accurate, if you have a very small dataset to process and you have lots of time. The second method is to use ica with MEGnet automated classification of artifact components. MEGnet was retrained on data from the CTF, Elekta/MEGIN, 4D, and KIT data. The model classifies components with >98% accuracy, so this is also an excellent option.
- What about Quality Assurance?
- Once all the processing is complete, you can generate QA images using prep_QA.py. Like process_meg.py, prep_QA.py will operate either on a single subject or on all subjects listed in a the .csv file produced by parse_bids.py. Once the .png files are created, you can use Run_enigma_QA_GUI.py to interactively label your subject images as good or bad.
- My Subject .csv file is not being read. Help!
- Check your file for errors.
- Does your path name have a leading space?
- Did Excel delete the leading zero from your run numbers?
- Does your path have any errors?
Main Processing Pipeline
Under construction
Don't see your question?
- If it is a script problem-
- Open an issue on GitHub.
- Otherwise-
- Email Anna Namyst and Jeff Stout