Okay, so let's start right away.

Where are we currently?

We talked about X-ray imaging.

Is there someone in the back who is recording?

Yes, here we go.

We talked about X-ray imaging

and acquisition-specific image enhancement.

We talked about image undistortion

and defect pixel interpolation.

Tomorrow morning I'm going to say a few words

about defect pixel interpolation,

what I've promised for last week already.

Now we will proceed and look into

our magnetic resonance imaging

and what type of artifacts show up

in magnetic resonance imaging

that have to be treated by algorithms.

Huge program for today and tomorrow.

First of all, I would like to revisit

the basic idea of magnetic resonance imaging

without going into the details.

And later on we will talk about

two artifacts that are crucial,

the so-called bias and gain fields

that make magnetic resonance images appear rather,

how should I say, unbalanced,

and they look very inhomogeneous,

and this has to be corrected.

Before these images are shown on the monitor.

But before we go into the discussion of the algorithms,

let's talk a little bit about MR imaging.

Here in this figure you see two typical MR systems

from the typical local company

that is building these systems.

These are two magnets.

And that's usually in clinical routine.

Today we have, let's say, 1.5 up to three Tesla systems.

In the hospital in Erlangen in Neuroradiology,

we have a new three Tesla system,

Magnetom, Vario, and the Trio scanner.

The Trio scanner is suitable for whole body scans.

And more research systems that are out today, they have up to

nine Tesla already.

So they have high field magnets,

and with these high field magnets, of course,

you get much better contrast and spatial resolution

than with these, but three Tesla is more or less

the state of the art in the hospital.

The high Tesla systems are not that common,

and I have to admit that we don't have any