The following content has been provided by the University of Erlangen-Nürnberg.

Okay, welcome to another lecture. Today the professor is not able to come, so I will give the talk.

For a few weeks you have started working on the analysis part of the biomedical signals.

Actually last week you were talking about filters in time domain and today we will finish that part.

But first let's start with the short test. So can you tell me low frequency sources or what could be possible sources of the low frequency artifacts?

Yes?

Yeah, for any, for some, especially for example for high frequency you can say this is high frequency for this signal and it's not for that, but now we are talking generally.

Do you remember in ECG, yeah?

Low frequency, the covering blinds?

What's the power?

Covering blinds.

Baseline wonder.

Yes.

Baseline wonder or motion artifacts.

That's when the electrode is moving. Okay, anything else?

It can be temperature drift also.

These are, let's say, related to equipment and do you know some physiological? Maybe EGG?

And for high frequency?

Yes, power line. When you are talking about, for example, EEG or ECG, then power line would be. For EMG it won't be because it will be in the same area. But yeah, we can write power line.

Okay, let's just say EEG or ECG.

And I just said one that is from physiological source.

EMG and ECG.

And of course there is some white noise that could be from recording system or amplifiers.

Okay.

I hope that you can read it. I'm not...

Okay, so the second question.

What are the fundamental requirements that need to be fulfilled in order to perform synchronized averaging?

Yeah, actually use your perception is that the your noise will have zero mean in a sense that yeah.

So about the noise.

Anything else?

What are you averaging? So what do you need to have?

Yes?

Yes, so about the signal. It's ergodic and it's multiple realization.

And you need to have a trigger or some alignment point.

Okay.

So this will be magnitude response.

Okay.

Okay.

Okay, and this will be the phase in degree.

And again here on the x axis, we have a frequency in hertz.

So when we are plotting magnitude response, the which frequency we are plotting it.

So which value will be here?

500 because we are sampling with one kilohertz. So it will be half of the frequency sampling frequency.

So here you will have 500 hertz.

And as for the magnitude, because we are plotting it in a dB, our starting point is going to be zero.

So do you remember the moving average is which type of the filter, which frequency low pass?

Yeah. So it will be something like this.

So here we will have around minus 100 dB and around minus 20 dB will be about 400.

But I don't think that actually when you're drawing the schematic, you need to know all this.

Let's leave it like that.