Page 1 of 1

Filtering and sensor selection

Posted: Fri Feb 10, 2023 9:32 am
by nlcg8
I was wondering if you clarify some questions!

1. is the data filtered before being saved in the CSV file?
2. In my experience connections on the temporal electrode sites is drops in and out. How is this taken into account when using the online graphing? If I just click and remove them, will the average be taken from just the front sensors?

Many thanks!

Re: Filtering and sensor selection

Posted: Fri Feb 10, 2023 9:36 am
by James
#1 - RAW EEG is unfiltered. Brainwaves are filtered for powerline noise and blink/jaw clench removal.
#2 - The graph shows the average of whatever you have selected. If you un-select an electrode then it will change the average displayed.

Re: Filtering and sensor selection

Posted: Fri Feb 10, 2023 11:51 am
by nlcg8
Thanks for the speedy reply!

Do you know what the filter settings are? I'm looking at using the muse headbands for a student project but as they are first year undergrads with little maths, I'm wondering if I just direct them to the online mind monitor graphing or if we should write some code for them.

Re: Filtering and sensor selection

Posted: Fri Feb 10, 2023 1:02 pm
by James
The filtering is done in Interaxon's closed source code. From what I know from talking to the Interaxon devs, it's band filtering out powerline noise at 50/60Hz and then dropping packets when it detects a blink or jaw clench.
The RAW EEG is untouched, so if you set your Recording Interval to "Constant" in settings you can do your own FFTs on that and calculate unfiltered waves.
If you've got students who you want to play with the data, check out https://mind-monitor.com/FAQ.php#oscspec where there is a link to my GitHub with Python code samples to work with the streaming data.