methlabUZH / automagic

Automagic
GNU General Public License v3.0
95 stars 32 forks source link

Error with channel location (BIDS dataset #71

Open eledau opened 6 months ago

eledau commented 6 months ago

Hello, I have been trying to pre process my data that are stored in BIDS format. These are the settings I am using: Screenshot 2024-05-13 172704

This is the error I get: Screenshot 2024-05-13 172942

In each subject folder I have a .tsv file (that is supposed to have the channel information), a .eeg file with raw eeg data, a .json file and .vhdr and .vmrk files. I have also tried to not provide a channel location document or create one myself and save it as .sfp but I keep getting the same error. I also tried to specify one of the subject's .tsv files as channel location file but it didn't work.

ksgfan commented 6 months ago

Hi,

it seems that the loading of chanlocs file failed.

if you already have tsv files with electrode locations in each subject folder, you don’t need to provide a new channel location file or file type. Automagic will find these tsv files on its own. Simply remove both entries.

Hope it helps.

Otherwise you could send me data of 1 subject and will figure it out for you.

eledau commented 6 months ago

I tried what you suggested but I keep getting the same error, how can I send you the data of 1 subject? Thank you for your help

ksgfan commented 6 months ago

You can use google drive, Dropbox or wetransfer. I will have a look at the data tomorrow.

eledau commented 6 months ago

https://drive.google.com/drive/folders/1Hl-91rN7LqsLUMGhqPED-6gRSE6Onuy6?usp=drive_link

ksgfan commented 6 months ago

thanks!

Could you also provide the '*electrodes.tsv' file or a channel location file that contains locations for 'VPVA', 'VNVB', etc. channels? Brainclinics Foundation

eledau commented 6 months ago

I do not have that file, but I copied can send you the channel location provided in the paper where the dataset was presented: https://drive.google.com/file/d/1QRsFDnVTCZq3nxiZ68IgnkUm6WU5ESQF/view?usp=drive_link After inspecting the .tsv file in my dataset i noticed that it does not contain channel location information, sorry!

ksgfan commented 6 months ago

First, please remove the first line from the .sfp file. This file should look like this:

Fp1     -26.81  84.06   -10.56
Fp2      29.41  83.74   -10.04
F7      -66.99  41.69   -15.96
F3      -48.05  51.87    39.87
Fz        0.90  57.01    66.36
F4       50.38  51.84    41.33
F8       68.71  41.16   -15.31
FC3     -58.83  21.02    54.82
FCz       0.57  24.63    87.63
FC4      60.29  21.16    55.58
T7      -83.36 -16.52   -12.65
C3      -65.57 -13.25    64.98
Cz        0.23 -11.28    99.81
C4       66.50 -12.80    65.11
T8       84.44 -16.65   -11.79
CP3     -65.51 -48.48    68.57
CPz      -0.42 -48.77    98.37
CP4      65.03 -48.35    68.57
P7_T5   -71.46 -75.17    -3.70
P3      -55.07 -80.11    59.44
Pz       -0.87 -82.23    82.43
P4       53.51 -80.13    59.40
P8_T6    71.10 -75.17    -3.69
O1      -28.98 -114.52    9.67
Oz       -1.41 -117.79   15.84
O2       26.89 -114.68    9.45
VPVA      
VNVB       
HOHL      
HNHR       
Erbs       
OrbOcc     
MASS 

Next, use these configurations: Bildschirmfoto 2024-05-14 um 10 03 29

You have to remove channels that don't have locations on the scalp (i.e., Exclude Misc. Channels). Otherwise, PREP and ICA won't work.

I was able to successfully preprocess both files with this settings:

Bildschirmfoto 2024-05-14 um 09 29 02

Let me know if it works for you.

eledau commented 6 months ago

Thanks! this solved my problem!