fMRI Analysis with SPM12: Functional Preprocessing (with GUI)

Поділитися
Вставка
  • Опубліковано 4 лис 2024

КОМЕНТАРІ • 14

  • @chauhan100990
    @chauhan100990 3 місяці тому +1

    Best video on UA-cam for fMRI preprocessing. Unfortunately, I found it very late. But, great job, great explanation.
    Thank you :)

    • @4Educ8ion
      @4Educ8ion  2 місяці тому

      Thank you very much, Chauhan! I hope the tutorials are still helpful for you, despite being a bit late : )

  • @salomonkhabie806
    @salomonkhabie806 Рік тому +1

    Hi, Thanks for your videos ! How should we adapt this when using multi-echo functional images ? Should we select them as being different sessions for each echo, and apply each step independantly ?

    • @4Educ8ion
      @4Educ8ion  Рік тому +1

      Hey Salomon, great question! Here is some useful info for handling multi-echo acquisitions: tedana.readthedocs.io/en/stable/multi-echo.html#processing-multi-echo-fmri
      So the typical procedure seems to be to estimate motion parameters from 1 echo only and apply that to all other echos. Slice timing should be performed before multi-echo denoising / combining, and the other steps (distortion correction, normalization, smoothing) after denoising.

  • @dheemant-dv3uz
    @dheemant-dv3uz Рік тому

    Hi, I am using the same pipeline for the HCP connectome data, preprocessed version, my function image seems zoomed in a bit and reduced in number of voxels, whereas the structural still stays he same, but contains all the areas of the brain when viewed in display mode. Is this a glitch or is expected out of it

  • @meenakshimalviya1042
    @meenakshimalviya1042 Рік тому +1

    Hi. I have applied the steps given in the tutorial as it is given, no change in any parameter. I had received the output files too, but my sswaurest.nii file is not displaying any result. when I display the image it shows a blank window, I has also tried to get the intensity using python but max and min both are 0. what does it mean? Do I need to change any parameter or any other suggestions please

    • @4Educ8ion
      @4Educ8ion  Рік тому

      Hi, thanks for your question! Did you make sure to select all images of the timeseries using 'Inf' (see 2:23)? It's a common mistake to only select the first image of the timeseries - it often happens to me too.

  • @mrthanioruvan4406
    @mrthanioruvan4406 Рік тому

    I want a dataset of fMRI (task : meditation ).If you dont have it,atleast tell me where can i find it?

    • @4Educ8ion
      @4Educ8ion  Рік тому

      Hi! You can find free fMRI datasets (including the one I'm using for this video series) on OpenNeuro: openneuro.org/

  • @jfdc4576
    @jfdc4576 5 місяців тому +1

    is this the prepocessing resting state fmri?

    • @4Educ8ion
      @4Educ8ion  5 місяців тому

      Good question! While my pipeline is generally based on task-based fMRI data in mind, the main preprocessing steps are essentially the same for resting-state data. So, I believe you could use this preprocessing pipeline also for resting-state. However, you may want to do more careful nuisance modeling later in first-level analysis, especially for motion artifacts (e.g. "motion scrubbing" based on framewise displacement) as well as physiological artifacts (e.g. aCompCor). While you can get these things yourself using Matlab code and toolboxes, the software fmriprep is a nice preprocessing pipeline that outputs a bunch of these nuisance regressors automatically: fmriprep.org/en/stable/ So this may be easier to use in the long run for resting-state data.

    • @jfdc4576
      @jfdc4576 5 місяців тому

      @@4Educ8ion I really appreciate your attention. It would be a great idea if you could make a video of this preprocessing for all students like me.

  • @clashwithvictor3365
    @clashwithvictor3365 Рік тому

    how can i aquire the data set that u have ?

    • @4Educ8ion
      @4Educ8ion  Рік тому

      The dataset is available on OpenNeuro: openneuro.org/datasets/ds000117