Which files to download from Openfmri Data?


#1

I am new to neuroscience and I am learning to work with neuroscience data.
However I am a bit confused. I tried to download the dataset:
https://openfmri.org/dataset/ds000116/
but somehow, I cannot understand which files to download, if I click Browse Data For All Revisions on S3 , then I can just find Anatomical and functional Images, but I am looking for both EEG and fMRI data.
If I download Raw data on AWS or Curated dataset its really small in size, as compared to Processed data for Subject 1 on AWS for each subject.

So my questions are:

  • What is the difference between Browse Data For All Revisions on S3 ,Raw data on AWS,Curated dataset and Processed data for Subject 1 on AWS for each subject?
  • Which dataset should I download?
  • Is the dataset Processed data for Subject 1 on AWS for each subject, is pre-processed, if yes then what pre-processing steps have been done?
  • can someone point me to resources for pre-processing EEG and fMRI data, can I use fmriprep for fMRI data?

I am a little confused why my posts are flagged?Can someone please suggest?


#2

What is the difference between Browse Data For All Revisions on S3 ,Raw data on AWS,Curated dataset and Processed data for Subject 1 on AWS for each subject?

Browse Data For All Revisions on S3 leads you to a browser where you should be able to find the other three versions of the dataset you mention (Raw data on AWS, Curated dataset and Processed data).

Which dataset should I download?

It seems that there are two revisions of this dataset. The first one (R1.0.0) is not in BIDS. The second (R2.0.0) is in BIDS but the EEG is not included. Since you want to have EEG, you should get all subjects from: https://openfmri.org/s3-browser/?prefix=ds000116/ds000116_R1.0.0/compressed/, which contains the compressed files corresponding to the revision 1.0.0.

Is the dataset Processed data for Subject 1 on AWS for each subject, is pre-processed, if yes then what pre-processing steps have been done?

All those links with preprocessed data contain the raw and preprocessed data as they were uploaded by the original submitter and briefly described in the README file https://openfmri.org/media/ds000116/README_8e6eXke.txt

can someone point me to resources for pre-processing EEG and fMRI data, can I use fmriprep for fMRI data?

fmriprep is definitely an option. I would first check on https://openneuro.org for this dataset to be preprocessed with fmriprep. I’ll repost here if I find it. If you are to run fmriprep yourself, then I’d recommend downloading the version R2.0.0 because R1.0.0 is not in BIDS format (and fmriprep will not accept that).

You could extract the EEG information from R1.0.0 and add it once you have the fMRI data pre-processed.


#3

Follow-up: currently, there is one “analysis” running fmriprep on these data at https://openneuro.org/datasets/ds001020/versions/00002 . If that analysis is successful, then you should be able to fetch the corresponding pre-processed data.

EDIT: in a previous snapshot of the dataset (https://openneuro.org/datasets/ds001020/versions/00001) you’ll find a successful run of FMRIPREP. You can even browse through the reports to check the quality of preprocessing (I haven’t checked myself). There is a DOWNLOAD ALL link if you want to get everything.


#4

Thank you very much oesteban for replying, really appreciate.
I was looking for the difference in meaning of different data sets: I see that raw and curated data are so small in size, around 5 GB.
But Processed ones are really big for each subject


Can you please suggest about below:

  • I want to understand the difference between: raw,curated, and processed data, lets say terminology wise, for example raw might mean that the recorded data without any preprocessing, if thats the case then the size of raw is really small, which is confusing. What does curated means.
  • In the dataset you suggested, what does the raw data contain?
  • I can see that on openneuro for my data set FMRIPREP - v9, had run successfully, so I can just download the processed fMRI files from this link?

#5

I want to understand the difference between: raw,curated, and processed data, lets say terminology wise, for example raw might mean that the recorded data without any preprocessing, if thats the case then the size of raw is really small, which is confusing. What does curated means.
In the dataset you suggested, what does the raw data contain?
I can see that on openneuro for my data set FMRIPREP - v9, had run successfully, so I can just download the processed fMRI files from this link?

Yes, “raw” refers to the recorded data without any preprocessing. Take a look here to see what pipeline we used for the “processed data”: https://openfmri.org/data-processing-stream/ . And yes, the processed data takes up much more space than just the raw data.
“Curated” means that we took the time to inspect the data, metadata, and organizational structure and correct any errors (or at least make a note of them).

In the dataset you suggested, what does the raw data contain?

I assume you’re referring to the openneuro dataset? This is the same dataset as ds000116 on openfmri. Openneuro is a site we are developing that will eventually replace openfmri. But it’s still in beta and the accession numbers don’t match between the two sites yet.

I can see that on openneuro for my data set FMRIPREP - v9, had run successfully, so I can just download the processed fMRI files from this link?

Yep!

Let me know if you have other questions.


#6

Hi jbwexler,

Thank you very much, really appreciate.

  • In the dataset you suggested, what does the raw data contain?
    I was asking about the raw data on AWS as is pointed out in the below pciture , Is it there just for reference or does it have some meaning?

  • Can you suggest some resources for a beginner in Neuroscience?