Hi, everyone.
I have a small doubt on hardware artifacts of EEG signal
I have a single sensor EEG data with a sampling rate of 1000Hz.
In one second-EEG data (1000 samples), if all samples have the same value
can I consider it as artifacts?.
If it considered as hardware artifacts, among 1000 samples how many
distinct values are required to believe that data is real EEG?
For example among 1000-samples of one-second data there is ten (10)
distinct amplitude can I considered it as hardware free artifacts ?.
There should always be some sensor noise at all frequencies. So, given a typical sensitivity of your analog to digital converter, I would regard even two consecutive values that are exactly the same a hardware artifact.
A good way to test whether you are recording actual data is to first record near the eyes and blink a few times. That is not EEG, but should give some nice peaks. Same goes for biting down, that should generate a huge signal due to EMG activity. Finally, to check for real EEG, measure at the back of the head with your eyes closed and all head and shoulder muscles as relaxed as possible. It varies from person to person, but likely you?ll see a smooth sine wave, around 10 Hz., starting to occur in the signal (your alpha wave).
Thank you vanvliet
For my work, I am using online downloaded EEG data.
among 1000-samples of one-second data, ten (10) distinct amplitude values
are sufficient to consider that data is real EEG data.
What is the minimum distinct amplitude values to consider that data is real
EEG data.?
[image: image.png]
In the above figure, each row is one-second data, and two values are
repeating every time.
Q1) What is the minimum distinct amplitude values to consider that data is
real EEG data.?
Q2) From the attached figure, can we consider such sensor data for further
analysis?
Q3) In some cases, few rows (one-second data) has same amplitude values
and few other rows as distinct amplitude values. In such a situation can I
neglect rows that contain the same amplitudes? [if I required one-second
epoch EEG data].
Q4) if Q3 problem is happening in case of the column, can we solve as like:
one sensor 5-seconds data with sampling rate=16Hz
each row is one-second epoch EEG data (5 rows --> 5 seconds data)
[image: image.png]
*if I neglect columns that have only zero values (amplitudes), then the
number of samples per second is 8 (shown in below table)*
[image: image.png]
in this case, can I consider sampling rate as 8Hz or still it is 16Hz
If you clarify my doubts, It would be a very big support for my research
work. Any time of your convenience is okay for me.
I am looking forward to a positive response from you.
THANK YOU SO MUCH FOR YOUR TIME AND CONSIDERATION