Hello, researchers. I am working on filtering an EEG signal with a FIR band-pass filter. I now need to define the cutoff frequencies of the band-pass filter, based on other previous data/studies that involve this.
I am having a bit of trouble doing that, so I am looking for a reference of something like this that's been done previously to be able to compare. Did any of you run into such work before, and would you be able to share it? Thank you!
Dear Iulia Turca , the choice of cut-off frequencies would depend on the subsequent methods one might want to use. For event-related potentials (ERPs), there is one crucial issue: if using a band-pass filter, the lower cut-off should not be higher than 0.1 Hz, as this might distort the ERPS (see the articles below). The higher cut-off is chosen below the first line noise harmonic, e.g. 40 Hz. For ERPs 0.1 - 40 Hz is thus a common choice. A nuanced discussion can also be found at https://sapienlabs.org/pitfalls-of-filtering-the-eeg-signal/
P.S. I apologize if this is nothing new, but for ERPs, Prof. Luck's book is one of the bedrocks, and discusses many pitfalls with a lot of detail: https://mitpress.mit.edu/books/introduction-event-related-potential-technique-second-edition
Article Systematic biases in early ERP and ERF components as a resul...
Article How inappropriate high-pass filters can produce artifactual ...
Hello,
Have you tried here?
https://brainclinics.com/neurofeedback-resources/
Hi, the selection of cutoff frequencies is highly dependent on your application. For instance, if you want to evaluate processes related to alpha rhythms, you should filter your data in the range of 8-12 Hz.
- Badges
- Science topic
- Similar topics
- Control Systems Engineering
- Filtering