My name is Tobias Knopp. I'm working at the University Hospital at Hamburg, Germany, where we have installed the first commercially available MPI scanner. We're very happy about that because this is really a step forward for MPI.
Up to now, we've seen a lot of improvements made, but now we're almost at the point of focusing on the application of MPI and proving what the real benefits of this technology are, which is very exciting.
My background is actually in computer science. I studied computer science in Lübeck in 2007 and for my PhD thesis, I started to work on MPI. We had a strong cooperation with the inventors of MPI (the Philips Group), which was really exciting because we were building MPI scanners and developing proofs of principle.
My background was focused on the image reconstruction part, which involves working out how to translate the raw data into an image. In MPI, the particle dynamics makes this very complicated, but we did also build scanners, so we also did core design.
I then went to Bruker and built the first preclinical MPI scanner. I wrote the software for that device and then got a call for professorship in Hamburg. Now, I am a customer for Bruker and I use this nice hardware.
The crucial thing about MPI is that, without any tracer material, we won't see anything apart from a black image. Therefore, we need magnetic nanoparticles, which are injected into the human or animal body so that we can trace where these particles go. This is crucial because it means we have a positive contrast for these particles, which is the really exciting bit about MPI.
MPI uses magnetic fields to spatially encode the nanoparticles and, in particular, a magnetic field that has a field-free point at the center is used. This field-free point acts as a sensitive spot and if we move this point around, we can get a signal, especially from the region surrounding the field-free point. We just move it around and scan the whole area along certain trajectories, which allows us to reconstruct a map of the particle concentration.
Of course, as a theory, this all sounds quite easy, but in practice, there are various challenges to face because particle physics is not that easy. There are various dynamic behaviors which are not so easy to predict and you have to get a lot of things right in order to obtain the images.
Image reconstruction is an important aspect of MPI. If the tracer goes into tissues and is immobilized, for example, it reacts a little bit differently. We have to compensate for that with an image reconstruction.
MPI is also a very fast technique and we get image data at a rate of over 40 frames per second. We have to not only reconstruct it, but we have to reconstruct it quickly because what we eventually want angiographic imaging; we want to observe flowing blood and for that, image reconstruction has to be really, really fast. This is also something which is computationally very intensive and therefore one has to write clever software to obtain the particle concentration.
MPI is actually new in the preclinical field and we are currently in the phase of exploring different applications with the scanner that we got from Bruker. We are moving in various different directions to find out how we can prove that MPI offers benefits over other modalities, which is crucial.
If we do MPI preclinical imaging, we always have to compare it to existing modalities to really show the benefit. We start with the cardiovascular system, observing and measuring blood flow in the cardiovascular system of mice and rats. We want to see perfusion, but it would be very nice if we could eventually reach a sensitivity level that allows us to do cell tracking, which is one of the major applications that MPI can hopefully be used for.
Until now, the situation has been that data is often acquired and reconstructed in an off-line step and it is a challenge to turn this into a one-step process where one can see what's going on directly, which is again crucial.
There are also other difficulties. Due to the dynamic behavior of these particles, we need a calibration, which takes a lot of time and is really challenging, although there are some good approaches to solving these issues. Very good models of the particle physics are available, which we can exploit in image reconstruction.
Today, at the conference, we have seen various new reconstruction techniques where colleagues have shown that one can do colorized MPI in order to distinguish different particles from each other, which is very exciting. We've seen a catheter inside of a blood vessel, where one was colored red and the other green. We are just at the beginning of exploring all these exciting possibilities.
My intentions are, especially in image reconstruction, to make it fully automatic because currently we adapt various parameters, which is not that straightforward.
Another very important aspect is sequences and we have the opportunity to design various sequences. We can make a low-resolution pre-snapshot and then progress to a higher resolution. However, there are many questions surrounding this, which is something I particularly want to investigate.
My intention is also that all these things should be validated in mice and rats in-vivo because we really have to now carry all of this over to the applications. This is what I will do with my group and scanner. I'm not a radiologist so I'm not an expert in that field, but I clearly see the cardiovascular applications as one potential application for MPI. Hopefully we will eventually get to the stage of cell tracking, so that we can actually see a tumor and that the particles went to the tumor, where we can detect them with MPI.
There are also some quite different applications such as monitoring, which is a big thing. One can monitor whether a patient has bleeding or something similar over a long time. We now have to prove that the various applications work.
My vision is that we see not just one clinical MPI scanner, but thousands introduced to clinics. Nowadays, we have such a variety of different modalities; we have MRI, CT, PET and it's really hard to come up with a new modality and say it is better. However, this is the vision and bringing this to the clinics would be great.