Welcome back to Beyond the Patterns. So today I have the great pleasure to announce Julia Schnabel,

a very well-known member of our Micae community. She graduated with a master in computer science

at the Technical University of Berlin and a PhD in computer science at the University College London

and subsequently held postdoctoral positions at University College London, King's College London

and University Medical Center Utrecht before becoming first associate professor and then

full professor of engineering science at the University of Oxford. In 2015 she joined King's

College London as chair in computational imaging. Julia's research focuses on machine and deep

learning, complex motion modeling as well as multimodal and quantitative imaging for a range

of medical imaging applications. She is serving on the editorial board of Medical Image Analysis,

is associate editor for IEEE Transactions on Medical Imaging and IEEE Transactions on

Biomedical Engineering and has recently founded the new free open source journal

of machine learning for biomedical imaging, the Melba journal. She has been program chair of Micae

2018, is a general chair of IPMI 2021 and will be the general chair of Micae 2024 to be held for the

first time in Africa. She is elected member of the IEEE EMBS administrative committee and the Micae

society board of directors and an elected fellow of the Micae society, Ellis and IEEE. So it's a

great pleasure to have her here and today she will give a presentation that is entitled smart

imaging from sensors to information. Julia it's a great pleasure to have you here and I'm very much

looking forward to your presentation so the stage is yours.

Thank you very much Andreas and thank you everyone for attending. I'd love to be there in person

and I love the area quite often actually in lovely Franconia. I've been there last summer

after isolating suitably just for a few days but of course there's no point in currently

coming for a lab visit if you're all at home anyway. So thank you very much for joining.

This is kind of the area where I would normally be working, central London in all its glory.

I'll see whether you can actually see my pointer. I'm not sure whether you can. London Eye is one of

the main landmarks now. St Thomas's hospital is this building over here and we're just opposite

now but we've got a lot of lab space in St Thomas's. So I'm going to go to the

lab space in St Thomas's and a fantastic view across the river to Westminster.

So today I want to talk about smart medical imaging. It's a bit like bringing

colds to Newcastle, not anymore since the miners stopped now doing colds in the UK but I mean I know

you've been really really active in that area so I feel quite humbled to actually be able to present

to you. Let me just get a bit more shade. It's very sunny in London today. I have no disclosures

other than the funders who support our work. One of the main funding sources we currently

have that I'm reporting on here is the Smart Heart grant which is a joint collaboration

with Imperial College London and Oxford and I find which again is with Imperial College

and us at King's. So these are the two main ones I'm going to report on but there's a lot of other

things going on at King's and I'm very grateful for that funding. So smart medical imaging.

In medical imaging we normally fall into one of these following groups which are almost silos.

We either work on image acquisition where we generate raw data and use some imaging sensor

or we work on image reconstructions where we want to transform the raw sensor data

into some image or other form of information that we can then view. Or I traditionally come

from the image so-called post-processing corner where we start with filtering reconstructed images,

we segment them, we register them and so on. And then the actual image analysis would start

where you try to make more sense of the data where you do volumetry, where you construct models,

where you start detecting things and maybe run a classification. And then comes the clinical

image interpretation. Usually this is like a pipeline of very disjoint modules that you're

targeting and it goes from medical physicists over computer scientists to clinicians.

Machine learning has kind of taken over our world. I mean it's been there for decades of course but

with more data and more computing time we actually started tackling all these individual silos

for some time. But we still do it not across these different groups so we usually either do one or

the other. And some of our work, and I'm sure you work as well, I know you work as well,