Okay, folks.

Let me just go and get the most recent set of slides.

And then we can start right away.

As I said, for the following program, we have two specials for you.

No, actually three specials.

We will first of all visit Siemens Medical in summer semester.

See the exhibition room over there.

And we will also get a chance to have a look into the factory where the MR systems are

built or CT, depends on where we are going to.

And the second highlight is that we will visit BrainLab in Munich for a day.

They build navigation systems.

So exactly what we are currently discussing, so hand-eye calibration, how you can compute

the position orientation of devices in space, how you can correlate it with preoperative

CT data, how you can overlay fade the CTOS where the surgeon is working on with CT.

CT image showing artificially your instruments and things like that.

So we will also visit BrainLab.

And besides that, we will also have a guest speaker two weeks from now, Professor Doerfler.

He is an interventional radiologist and he is using interventional imaging techniques

in his daily work and he will explain to you a little bit what his life or how his daily

life is, how interventional imaging works in practice and what his demands with respect

to algorithms are basically.

And the reason why I am that late is because I just had a meeting with him and we were

discussing future collaborations because this type of interdisciplinary research is something

that we all appreciate a lot.

So welcome to the Tuesday session and we are in the middle of one very interesting chapter

that deals with the so-called hand-eye calibration problem.

And the figure you have to keep in mind is that we have our, I have one figure here,

hold on a second.

We have our ultrasound probe, here you see the ultrasound image you can get, you have

your patient, here you have your markers and you can have for instance preoperative MR

dataset or CT dataset and you have a tracking system that is basically observing the scenario.

That's the big picture.

And what we are working on today is how can we compute the transformation between the

image coordinate system and our marker coordinate system.

So the hand and eye transformation is something that we have to compute, that we have to compute

and we have to estimate using computer vision techniques.

And this is very important, I mean just mentally replace here the ultrasound probe by an endoscope.

Then you have a lengthy endoscope like this one here, a laparoscope where you have your

lens system, here up front, here you have your markers and you move the whole thing

around in 3D and you want to know how the coordinate system of your markers is related

to the coordinate system of your image.

You have to know that if you want to augment for instance CT dataset or MR datasets with

your devices that you are currently using.

So that's a very important thing.

It might appear a little complicated to you if you hear it the first time, but on the

other hand you should have to keep in mind, look this is a very practical, important problem

of highest relevance for surgeons and treating physicians.

So scenario, as follows, we have here the endoscope, we have here the ultrasound probe.

You can also think about a knife of a surgeon or a pair of scissors or additional devices.

You can all put or extend these systems with markers and then you can compute their 3D