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

Good afternoon everybody to the Monday afternoon session.

We stopped last time discussing X-ray detector technology and the implications in terms of image quality

and how we can increase the image quality by using algorithms to correct for them.

I have to apologize, I screwed up the convolution of the 1D symmetric signal last time,

but this is not a big deal, but as I said, I screwed it up,

and you can look this up in the handouts we put on the web.

So today we will start the next chapter on magnetic resonance imaging.

And magnetic resonance imaging is another imaging modality that is becoming more and more applications in the medical field,

because MR, magnetic resonance imaging, has clear advantages compared to modalities like X-ray or CT.

MR is not affecting the human body like we know it from X-ray systems.

And before we look into algorithms, I will make a few historical remarks.

I will give you a rough, a very rough idea how MR works,

and then we will look into standard preprocessing methods that are basically implemented in today's MR scanners.

And most of the users are not aware of this preprocessing step.

So let's start out with the historical remarks.

So the first insights in magnetic resonance imaging grew up or grew in the 30s already,

where Rabi found a new method how to generate molecular magnetic resonance.

And basically what they were able to do, they were able to show molecules and to visualize molecules in a magnetic field.

In 46, and this was a very important stage, Bloch and Purcell, I think that's the way it's pronounced,

they discovered the nuclear magnetic resonance phenomenon.

That's something on a microscopic level where you basically can show that, for instance,

hydrogen atoms have some magnetic properties and that they have their own spins,

and that you can use this as a physical property.

And as usual in medical imaging and medical image processing, we use principles from physics to visualize things.

And the magnetic behavior of material and the way, or a particular way to visualize the magnetic distribution in the human body

is basically the core idea of magnetic resonance imaging devices.

And we'll not go through all the line here, and you should not learn these things, because that's history.

We don't need to know about this.

Interesting, in the 60s Anderson and Ernst, they introduced Fourier methods to enhance the magnetic resonance of nuclears.

And then at the beginning of the 70s, people started to look into the diagnostic usage of the nuclear magnetic resonance properties.

And that's basically the same timeline as CT.

At the beginning of the 70s, they started to understand how you can visualize magnetic properties in the human body

and how to use the magnetic properties to distinguish between different soft tissue classes and different materials.

And there are a few names you should have heard of.

One is DeMadian.

He basically initiated the medical application of magnetic resonance imaging by observing differences

in the nuclear magnetic resonance signals between cancerous tumors and normal tissues.

So he found out that normal tissue and cancerous tissue show a different behavior in terms of its nuclear magnetic resonance properties.

And in 1972, Lauterbuer and Mansfield, these are the two most important names in this context,

they developed a method to do a reconstruction out of the magnetic properties that are measured.

And Mansfield, he basically built the first MRI scanner and was able to generate an image of the human body.

Remember, CT was done in 1972, round about that, and that's exactly the same time range.

And Lauterbuer also borrowed algorithmic developments from CT and applied it to MRI reconstruction,

because it's basically the same technology that is actually used.

And we will have a whole section on image reconstruction methods, and then you will more and more understand how these parallels are.

And then, as in CT, the whole industry started right away to build these scanners and to use these scanners for practical use.

And it was nice in 2003 when I was teaching MR, the Nobel Prize was awarded to Mansfield and Lauterbuer.

And it was interesting at that time to read the press, because Raymond de Maidian,

he was very angry about the fact that he did not get the Nobel Prize.