From: University Nemiear on
Can anyone else tell me how to filter EEG signal using matlab? I am a newbie to this software.

Can we use fdatool?
From: Walter Roberson on
University Nemiear wrote:
> Can anyone else tell me how to filter EEG signal using matlab? I am a
> newbie to this software.
>
> Can we use fdatool?

You have a set of digital data, you need to filter it, fdatool is a tool
for constructing filters, ergo you can use fdatool to construct a filter
for your data.

It would be even easier to use one of the command line filter
construction tools and feed it random coefficients: if you don't know
what kind of filter is *appropriate* for your situation then you might
as well make that explicit by using a random filter.

Perhaps you should read some papers about the expected shape of your
signal, and about the possible noise sources and how they would affect
that shape, and only *then* start thinking about what specific tools you
might use to remove those kinds of noise without destroying your signal.
From: Wayne King on
"University Nemiear" <grin_face(a)hotmail.com> wrote in message <i3fln0$d61$1(a)fred.mathworks.com>...
> Can anyone else tell me how to filter EEG signal using matlab? I am a newbie to this software.
>
> Can we use fdatool?

Hi, As Walter correctly states fdatool is a tool for designing filters. If you feel like you have designed an appropriate filter using fdatool, then you can export your filter to the MATLAB workspace with File -> Export. You have the option to export either the filter coefficients or a filter object. If you export your filter as a filter object, Hd, you can filter your data with:

Output = filter(Hd,Input);

Note that this is not a zero-phase filter implementation and features in the filtered data will be delayed with respect to features in the original signal. (provided of course they have not been removed by filtering)

You can also generate the MATLAB code that produces your filter object by File -> Generate MATLAB code (In your version it may read File -> Generate M file).

If you wish to use zero-phase filtering so that features in your filtered data line up with features in the original data, then you will want to use filtfilt(). If you want to use filtfilt(), write back because there are a few details that you will need to know. filtfilt() does not currently accept filter objects as an input, so you will need to either export the filter coefficients from fdatool, or extract them from the filter object. If you design an FIR filter, this will be straightforward, if your filter is an IIR filter (Butterworth for example), then it's a bit trickier.

Hope that helps,
Wayne
From: Wayne King on
"University Nemiear" <grin_face(a)hotmail.com> wrote in message <i3fldj$ojc$1(a)fred.mathworks.com>...
> Hi. I am a newbie to Matlab. Can anyone tell me how to filter EEG signal using matlab?
> can we used fdatool?
see the thread in the other post
 | 
Pages: 1
Prev: filtering EEG signal
Next: Band power of EEG