Failure of recognizing 32-bit NeuroScan data

[Vladimir Litvak]

Dear Dan,

On Wed, Apr 7, 2010 at 2:59 PM, Dan Zhang <zd8472 at gmail.com> wrote:

> I've encountered a problem of loading the 32-bit NeuroScan data into
> FieldTrip.
>
> After one day debugging, I found out that the data was wrongly recognized as
> 16bit, which made all the things in the wrong place. I manually made some
> changes in several m-files to make it work for my current dataset by
> assigning all the processing in a 32-bit way, which cannot be the final
> solution.
> Anyone tell me how to automatically recognize the version of the data? Then
> I can try to fix this bug.

This is not a bug but a known issue. Until now we have not found any
way to automatically distinguish between 32-bit and 16-bit Neuroscan
data. This problem is also not solved in EEGLAB where the reader
originates and the user has to specify it via the GUI. What you can do
in your code that will not require modifying Fieldtrip code is specify
in configuration of ft_preprocessing:

cfg.headerformat = 'ns_cnt32';
cfg.dataformat = 'ns_cnt32';

Similarly if you use read_header, read_data or read_event there is an
optional input argument for data format that you can use. If you come
up with a way to distinguish automatically the two format variants
we'd be happy to hear about it.

>
> Yet here is another small problem: the imported NeuroScan data may not be
> with the correct unit (e.g. uV). I know the loadcnt() function can be evoked
> with 'scale' = 'on', but it seemed that this parameter was not used in the
> current version of FiledTrip.
>

'scale' = 'on' is default in loadcnt so there is no need to set it
explicitly in Fieldtrip functions.

Best,

Vladimir

----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/neuroimaging/fieldtrip.