Hi,
I was wondering about best practices for processing and especially pre-processing the raw EDA signal, particularly with AcqKnowledge, when solely interested in the Skin Conductance Level (i.e. tonic EDA). I'm looking for changes in arousal over longish periods of time 30-60 s and sometimes the level of (high frequency) noise in the raw signal is just crazy. Since I'm new with SCL, I'm not sure of what filters ought to be applied. Anybody got any suggestions? Would smoothing be more appropriate?
Thanks
I'm not a direct expert in pre-processing EDA signals (although I do work with other electrophysiology signals), but my suggestion would be to take a look at the following guideline paper: http://onlinelibrary.wiley.com/doi/10.1111/j.1469-8986.2012.01384.x/epdf
Section 3 provides a relatively detailed look into signal evaluation and parameterisation of a recorded EDA signal.
You can get useful information on analyzing EDA data using AcqKnowledge via BIOPAC webinars: https://www.biopac.com/webinars/
Any case anyone else is still wondering about this, what a colleague suggested to do was to use a 0.05 Hz low pass filter to get rid of any noise and phasic responses and then extract the mean.
I would suggest downsampling the signal to max of 60Hz, then using a moving window filter to remove noise (Typically, EDA value wouldn't increase by more than 20% or decrease by more than 10% in a 1 second window). Also, if there value is zero or very close to zero (
- Badges
- Science topic
- Similar topics
- Psychology