- About Braintrip
- Uploading/downloading a file using a terminal window
- Using the upload script
- Using the download script
- EEG recordings access
- Additional notes
The BrainTrip project is attempting to use quantitative analysis of EEG data to predict cognitive decline and dementia in seniors. We've performed several hundred EEG recordings using a 24-channel EEG amplifier and saline sponge based EEG caps. Our EEG data is encoded in the BrainVision format. Each EEG recoring has 3 associated files: a binary raw EEG file (.egg), and two ascii files containing marker (.vmrk) and header (.vhdr) information about special events in the EEG (markers) and channel names and positions (header).
Further information on BrainVision format are available here
To prepare your EEG data for upload, please convert your raw files to the BrainVison format for ease of use across different labs. Then follow the upload instructions.
There is no requirements in order to upload or download files by using the bee gateway. However, if you are using it for the first time, you need to create a user using a command provided in list of commands. Otherwise, you only need to log in by typing a login command in the terminal window.
In order to upload a file you need to pass an upload command to a terminal window (shown below). A file should consist a header file, marker file and raw eeg, archived using 7zip software to a .zip file.
If the upload is successful you should get a response similar to this:
HTTP/1.1 201 Created
Server: nginx
Date: Thu, 27 Oct 2022 07:33:45 GMT
Content-Type: application/json; charset=utf-8
Content-Length: 82
Connection: keep-alive
x-powered-by: Express
access-control-expose-headers: Swarm-Tag
etag: "469948072a5b62f382911293888fe960cb1e18160941282348e6f321b2aa076e"
swarm-tag: 567568429
swarm-server-id: gw_bee_be1
Access-Control-Allow-Methods: POST
Access-Control-Allow-Headers: *
The response consists of information about upload status, but the most important bit of information is the etag. This is a hash that was given to the uploaded file and is used in download command to access said file.
In order for other people to be able to access uploaded files, the etag should also be made available in this repo. It would be best to upload a csv table containing three rows: file name, timestamp, etag. The csv file should be named like this:
fileType + '_etag_' + datetime.now().strftime('%Y-%m-%d_%H-%M-%S') + '.csv'.
Example: eegRecordings_etags_2022-10-04_10-23-24.csv
To download a specific EEG recording, you need to log in and move to a desired directory via terminal window and execute a download command. The file should be downloaded in current directory under a specified name, which you write at the end of download command.
-
curl 'https://fairosfairdatasociety.org/v1/user/signup' -H 'Content-Type: application/json' -d '{"user_name":"brainTrip","password":"verySafePassword"}' -
curl 'https://fairos.fairdatasociety.org/v1/user/login' -H 'Content-Type: application/json' -d '{"user_name":"brainTrip","password":"verySafePassword"}' -
curl -i -X POST -H "Content-Type: application/zip" -T "./testEEG.zip" 'https://gateway.fairdatasociety.org.bzz' -
curl -X GET 'https://gateway.fairdatasociety.org/bzz/insert_your_etag_between these_backslashes/' --output fileName.zip
In order to upload files more easily, we wrote a simple upload script (upload_dir.py), which uploads specified type of files from the chosen directory to Swarm.
Firstly, you need to prepare the files, which are intended for upload and put them in a desired directory. Prepare the eeg recordings in BrainVision format (header file, marker file and raw EEG) and archive them using 7-zip software to a .zip file. Then run upload_dir.py script which takes 4 arguments:
- path to the directory where the .zip files are,
- file type, which has the information what kind of files we want to upload and in which format (e.g. eegRecording/zip --> note: there can be other files in the folder, but the script will pick out only the ones with extension .zip)
If the upload is successful, the script will generate a csv table, named by a formula: "fileType_etag_timestamp.csv". It has three columns; file name, timestamp, etag. The table is intended to give other users access to uploaded recordings. The csv file should be uploaded to this repo in EEG_recordings folder.
python upload_dir.py C:\path\to\upload\folder eegRecording/zip
If you haven't already it is advised to clone this repository for the download script to work properly. The download script takes three arguments:
- folder name, which will be created if need be,
- names of .csv tables separated with a comma or '-af' if you want to download from all of the tabels
- indexes of rows of .csv tables separated with a comma or '-ai' fo you want to download all of the files in a table (if there is 10 files in a table, you can use indexes fromo 0 to 9).
Below are some of the axamples of download script being used.
-
Download all of the currently uploaded 500 recordings:
python download_recordings.py folder_name -af -ai -
Download from only two .csv tables:
python download_recordings.py new_folder eegRecording_etags_2022-11-22_13-08-25.csv,eegRecording_etags_2022-11-22_12-31-17.csv -ai -
Download only certain files from a certain table:
python download_recordings.py new_folder eegRecording_etags_2022-11-22_13-08-25.csv 1,3,5
Tables of etags of uploaded EEG recordings can be found here.
You can also download or upload the files to swarm via swarm gateway online using this link. Please note, that if you download files from .csv tables via this link, they will be named after their hash. You can still use the download script to download recordings in bulk.