I seem to be unable to 'review raw file', & instead get the error message as shown in the attached screen-shot. I've followed all the (excellent) tutorials plus the EEG/epilepsy one; I'm using the Runtime compiler 2015b; & my computer has plenty of memory.
The steps I follow are: create a new protocol, with 'Yes, use protocol's default anatomy'; create a new subject; right-click on subject & select 'review raw file'. And get the error message.
The data was acquired using Neuroscan's Curry software & has not been pre-processed at all. Curry produces a .dat file. I've also tried replacing the default anatomy with the .rs3 file provided by Curry, but this gives the same result.
Is there something I need to do to the data before I attempt to import it? I'm sure it's something very simple! Any guidance would be very gratefully received
Hello,
The .dat files can be multiple things in Curry, the same extension is used for averaged ERP and for the positions of the electrodes. The extension .rs3 is only for 3D positions, so you are maybe trying to import the wrong type of data into Brainstorm.
If you are using the menu “Review raw file”, it’s probably because you are trying to read the continuous recordings from the Neuroscan software. In this case, you should be selecting the file format “EEG: Neuroscan”, and select a .cnt file.
Francois
Apologies for the delay in responding to your advice - I’ve been trying to master Brainstorm & really struggling with it.
The Neuroscan .dat files I’m working with are the continuous recordings. This is the format used by the CURRY software. I tried the .rs3 file only because I wasn’t getting anywhere with the Default Anatomy. Each .dat recording comes with its own .ceo, .dap, & .rs3 files, which is why I thought it might help.
For Review Raw File, the list of file types includes .dat files from Neuroscan. But that hasn’t worked with the raw data in my case, & hasn’t worked for a .dat file that was recorded using Neuroscan SCAN software in 2009. What kind of Neuroscan .dat files is Brainstorm designed to work with? I’m thinking that maybe I should be carrying out preprocessing in CURRY and transferring a later version of a .dat. Would that give better results, do you think?
Anyway, I have converted a .dat file to a .cnt file & am working with that - though still I keep encountering problems that aren’t addressed in the tutorials. For example, I’m working through an EEG analysis in tandem with the MEG YouTube video by Sylvain. After application of a notch filter & then a bandpass filter, I was unable to open the time-series for either of them. I found that I was able to resolve this problem by modifying the channel file. Is that a legitimate solution, or might it have repercussions for the data that I’m not aware of?
Many thanks for your last response, & I hope you don’t mind casting some further light on these issues.
The type of .dat files I know from SCAN are ASCII export (text files), not the original continuous recordings (which always have the extension .cnt). Working with text files is not recommended, especially for continuous recodings, because they are very slow to read and inefficient in terms of storage. Open the file with a text editor to verify if it text or binary data.
If the file was processed with Curry, maybe it created another type of .dat file… In this case we may need to add the support for these files in Brainstorm.
Send me one of these .dat files if you want me to look at it (upload it somewhere, eg. dropbox or google drive, and post the link here or send it by private message).
Ah. Thanks for that. Opening the .dat file in Notepad produces a long sequence of alphanumeric & symbol gibberish, so it’s not a text file.
I’ve uploaded a set of CURRY files to Google Drive and sent you an invite to access them. I trust that works okay as it’s the first time I’ve used Drive.
Thanks for the file.
This corresponds apparently to a Curry “Raw Float Format (.dat)” format, which is not supported by Brainstorm, but it looks like EEGLAB can read them so we can probably use the EEGLAB plugin as well.
I hope I will find some time to work on it in the next few weeks.
In the meantime, you can export again your files in a different format from Curry (Neuroscan .cnt or EDF+, both supported by Brainstorm).
That’s good to know, Francois. Thank you for looking at the file, and it would be wonderful if you’re able to accommodate the Curry .dat format in Brainstorm.
I’ve just worked on this. I’ve added code to read the file you sent me, but could not allow for all the combinations of options. At the moment, the files that are already epoched (as opposed to continuous files), the files saved in a multiplexed format (saved sample by sample instead of channel by channel), and the ASCII exports cannot be read.
If you have examples for these types of file, please share them with me.
Please update Brainstorm, link your recordings to the database with the file format “EEG: Curry”, and check carefully that all the information is correct:
Apologies for revisiting this issue after so long. I've been away for a while and am just back to Brainstorm . . .
Sad to say, I still am unable to open the Curry .dat continuous files (no preprocessing). If I right-click, select EEG, select Display Time Series, the program tries to Load Data but just keeps going and going without getting anywhere.
If I double-click on the Link to Raw File, I get an error message as attached. This is probably not a recognised way of proceeding, but I thought the error message might be informative with regards to the .dat file!
@yellamsettya The error message you are reporting does not match the current version of the Brainstorm code. Brainstorm was not updated properly, or updated before the fix was actually pushed to the server.
Update Brainstorm again. If you still get an error, make sure you don’t have multiple versions of Brainstorm in your Matlab path. For a clean install: delete all your brainstorm3 folders, and download a new one from the website.
