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

Welcome to today's lecture. We're going to talk about waveform analysis and instead of a short test,

and since I haven't seen you for a long time, so I want to bring you up to speed again where we are and myself as well,

we'll do a little summary of where we are.

So we talked about generation of biomedical signals in the beginning.

And you could also say that we are somehow observing, I'm a really bad drawer, but I would say you're somehow observing the world.

So some signals are generated and these signals need to be interpreted.

And we are specifically looking at stuff that's interesting for us, so probably humans that are part of this world.

So we have some stuff happening.

We observe this with sensors

and want to make use of these sensor data.

So we somehow need to find methods to pre-process this data

and then to analyze them.

And finally, we do this to draw or to make decisions, for example.

So what I just put on the blackboard is kind of the story of our lecture and it's also the pipeline of machine learning.

It's nothing else.

So in machine learning, you have some stuff happening, like for example, self-driving cars, they're driving on our autobahns.

And you need some kind of sensing principle.

So in self-driving cars, it's a camera, for example.

You pre-process the data, so you extract some edges from the camera image or some contours.

You analyze the data, so for example, where is a feature in the data that tells me, or in the video image,

that tells me there might be a traffic sign or another car or whatever, or a fire truck, which I then don't see.

And then you make decisions. I have to brake, I have to accelerate, I have to observe the speed limit.

Now in the context of our lecture, it's the very same pipeline, and by the way, the story of the lecture.

So we talked about the human first. That's what's happening, right?

So generation of signals.

Oops. Generation part one.

Because when you want to build a self-driving car, you have to understand very well where this car is driving.

Otherwise, you don't see the fire truck.

And by the way, this was in the news, I think, yesterday or the day before yesterday,

that another Tesla hit another fire truck on the autobahn because the car didn't see it or whatever.

So you have to know very well in what environment you're driving,

because you might see sheep on the road or you might see pathologic changes in the ECG of a patient.

And that's why we talked long and intensively about what signals are being generated in the human body.

Then we talked a little bit about measurement.

I know this was just in brief and not very detailed, but here you got some basic understanding of how we acquire data.

So we talked about accelerometers, which are inertial measurement units.

We talked about electrodes. We talked about some other things to actually get digital data into a machine that we can process.

So this was our second part of the lecture.

And for all these, so you could hear, get more in medicine.

So whenever I do research, I work very intensively together with medical experts,

because I want to have their expert knowledge in my system.

So today I had three different meetings before this lecture in three different clinics of our university hospital,

just to make sure that the expertise of the clinical experts is flowing to our research.

And here you could go to electrical engineering.

So for example, the lecture on medical electronics to get more.

So there's a lot of stuff available at this university.

Now, after the measurement part, we started our third part of the lecture.

So this is all the third part.

And that's, of course, part three.