search

BTajini / augfmri

Python implementation of Functional Magnetic Resonance Imaging data augmentation through conditional ICA (MICCAI2021-Oral)
BSD 3-Clause "New" or "Revised" License
3 stars 2 forks source link

Inquiry about Data Augmentation Toolbox for fMRI Images #5

Open MehdiMatinfar opened 7 months ago

MehdiMatinfar commented 7 months ago

Dear Mr. Badr Tajini,

I hope this message finds you well.

I'm writing to you today with great interest in the data augmentation toolbox you've developed for fMRI images. I'm currently working with my own fMRI dataset and am eager to explore the potential of your toolbox for augmenting it.

To ensure I proceed correctly, I'd be very grateful if you could clarify a few points:

  1. Dataset Compatibility:

Would it be possible to use my own fMRI dataset with your toolbox, or is it specifically designed for a particular dataset or format? If compatibility is possible, could you please provide any guidance on the necessary data preparation steps?

  1. A_rest.py File:

I've noticed a file named "A_rest.py" in the toolbox, referred to as the "A matrix." Could you please explain the purpose of this file and its role in the augmentation process? Is this file specific to a particular dataset, or can it be used or generated for other datasets as well? If it's dataset-specific, could you please provide instructions on how I might generate a similar file for my own data? Thank you so much for your time and consideration. I'm very excited about the potential of your toolbox and look forward to your insights.

bthirion commented 7 months ago

Dear Mehdi,

  1. There is nothing that prevents application to other datasets. My best advice would be to apply fmriprep pipeline for preprocessing, and thus use Nilearn to project the data into Difumo representation .
  2. Regarding A_rest, Idon't remember well how it was obtained. Reading the example code, I guess it is doable to use it as-is. Please let us know if it works or not. Best,