[Eeglablist] applying CleanLine results to unfiltered data

[Andreas Widmann]

Hi Raquel,

both high-pass filtering and cleanline are linear operations. You just need two subtractions. There are two possible strategies which should end up in the identical result.

(1) Subtracting the high-pass filtered and "clean-lined" data from the high-pass only data gives you the line noise estimate. You can now subtract the line noise estimate from the unfiltered data.
or
(2) Subtracting the high-pass filtered data from the unfiltered data gives you the low frequency part of the data. You can then add the low frequency part back to the high-pass filtered and "clean-lined" data.

Make sure to work with double precision data as recommended in the subsequent paragraph of the quoted ref.

Hope this helps! Best,
Andreas

> Am 01.09.2016 um 20:59 schrieb Raquel London <raquellondon at gmail.com>:
> 
> Dear eeglab list,
> 
> I am interested in using CleanLine to get rid of 50 Hz line noise. However, as I understand from the information I could find, it is important to use a ~1 Hz highpass filter on the data before applying the CleanLine algorithm. I would prefer to use the method described in the PREP pipeline: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4471356/
> 
> "We obtain essentially the same results if we high-pass the original signal, remove line noise from the high-passed signal, and capture the signal that was removed. We subtract the captured noise signal from the original signal to obtain a “cleaned” unfiltered signal. If we subsequently high-pass filter the cleaned unfiltered signal, we find the line noise has been removed as though the signal had been filtered prior to line noise removal. The result is similar for a 1 Hz high-pass filter. The PREP pipeline uses this strategy for its line noise removal to avoid committing to a filtering strategy for the final pipeline output."
> 
> However, and I apologize if this is a stupid question.., I am not able to figure out how to implement this strategy. Is there any variable or set of variables I can save after running CleanLine that I can then apply to the unfiltered data, similar to what you can do with ICA weights?
> 
> Thank you!
> 
> Raquel
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to eeglablist-request at sccn.ucsd.edu