So good morning everybody.

Tuesday morning session, we will talk about model-based catheter tracking today again,

but before we dig into the details and continue what we started yesterday, on Tuesdays we

always have the big picture, so let me just take a few minutes to summarize what we are

doing here and what we did so far and what the storyline of the whole lecture is, okay?

So just the mind map as usual at the beginning of the lecture, we have here the cloud saying

IMIP, Interventional Medical Image Processing.

It has to be fast, it has to be reliable, it has to be robust and so on, easy to use

and so on.

So we are taking care of systems that are basically used by the doctor during a surgery

during an intervention.

And at the beginning of the lecture, we started to look at magnetic navigation.

In the first week when I was not here, we did some pre-processing stuff that is important,

but basically the core topics we started in the second week and one topic there was magnetic

navigation.

Magnetic navigation.

So what was the idea of magnetic navigation?

The idea was that our catheter tip holds a little magnet, we have huge magnet on the

left and the right side of the patient.

We have a user interface that allows the doctor to tell the system this is the direction of

the magnetic field and then the magnetic field is applied to the catheter tip and by pushing

the catheter a little bit, it's going into the right direction.

The support the doctor gets with this type of device is it's much easier to navigate

through complex vessel structures like the cerebral vessel system or like the coronary

arteries in the heart.

So we have magnetic tools that allow us to change the orientation of the catheter and

to simplify the interventional procedure.

And the problem that we have considered was design a user interface.

And the design of the user interface must be done in a way that during the procedure

it's easy to adjust the magnetic field.

In an interventional suite, usually you have a C-arm system, we all know what a C-arm system

is with a detector and an x-ray tube and you can capture 2D projections.

The question that we tackled was how can I use the 2D projections to adjust the 3D orientation

of the magnetic field?

And the idea was that we captured two images.

We have here a 3D point, here our optical centers and out of this, oh I cancelled the

wrong, it doesn't matter.

We say this is the point and this is the point in the 2D image and this is the 3D point and

now we define two points here in this image and here in this image and we compute basically,

oh I'm not good in this, oh wow.

Let me just take this away here and bring this in again.

So we have here the point and here and here the vector and out of this we can compute

the 3D displacement vector between the 2D points.

That was the idea that we use 2D projections and mark two points in one image and two points

in the other image and then we say in the projections this is exactly the orientation

of the magnetic field and then we go back into 3D and recompute the three-dimensional

orientation vector and this is given to the magnets.

That was the idea.

And now we have discussed at the beginning, the more complex case by the way, the more

complex case that we can arbitrarily move around the C arm.