 So, thanks for coming here. I mean, first of all, I would say I'm definitely not an expert on simple terms. I mean, that's some part of this because the lesson to learn your ideas, what could we use, you're always open for interesting opportunities. Let's see how it goes come out. My name is Lake, and I work in Gatheburg for AstraZeneca, and there's an even imaging science team, not huge, but we work together, of course, we are colleagues in Cambridge. So, you know, it's the most expensive building in Britain ever. And then we are in Gatheburg, it's the whole annual that your anti-bodies, and there's a Boston side, there's more infection. And so just for Alex John, it's more into rare diseases. I think in the, it's normal, obviously, not a big company. AstraZeneca only care about mega brands, you know, but it turns out there's a lot of money actually to own even in rare diseases. And sometimes actually the rare diseases are quite of interest for an imaging perspective because it's very normal, it's a very known gene genetic modification that needs and that's very extreme anatomical changes. So imaging is I think it's very crucial also for very much for those rare diseases. That's the one? No, no. Okay, so you all know that the imaging is done every day, it's done in hospitals for diagnostic purposes, to follow diseases and to study the cure of diseases every day. And in other clinical trials, as the coroner said, in a phase one, in a phase three study, typically cost 10 billion, and it costs include 10,000 people. So you need to know basically before you get in a clinical trial, that you have sign of life, a leap of hope, such that they really strongly believe they see something. If you run an IPF study with prime fibrosis, you need loads of patients for a long time. So now there's a lot of interest, of course, develop biomarkers that say you can, like emphysema and somatic iteration, it takes again probably a couple of years if you see the effect on getting read out. And of course, imaging is very important for early decision or for the risk in late stage. And also, of course, it is really to differentiate from other compactors. It could be that you have a drug for asthma, but there are other drugs for asthma. But if you can then show that this drug also have an impact on emphysema, although that is not your primary read out for your clinical. So imaging is used for inclusion criteria in the diagnosis to kind of find the right patients for the right dose, the right compound. And my typical is of course oncology or IPF or some of the respiratory diseases where of course imaging is always established and you just use imaging to find the right subpopulation that you go for. It's always been argued that one of the main failure previously for a big company that you are. They try to think a drug work for all patients, which obviously doesn't. So the final right place. Sometimes we use the not very often as companion diagnosis companion to not diagnose is very. I would say very hard to basically you use something by a market to say, this drug only work with this subset of patients. And then basically you file that together with your drug into FDA. So if you have it. If you have run your study in certain disease population. And of course that comes even more into radicis that we if you want to scan for part failure when they are you know, we may be thousands of patients you obviously don't want to. So, so there is of course and a very attractive to do the same things pretty nice if you can run the same modality in a mouse that you do in a man, of course that's very attractive very something very that you like to do. And of course we can we can scan mouse off similar to her to a human are using just another MRI scanner, which mean high field and high strain. And so without to my mind my personal view is that that's great. And that's always a selling point for a machine, but I would say, at the end of the day is that each animal study has to answer a specific project was specific decision transit to say so it's very often that we use a completely different image modality for example, pre clinic that we use in clinic. But I could come back to that. But I can say for example in. So what we are working in is spec. We do an ultrasound. There's a lot of ophthalmology by myself. There are some pros, some like another car in the show. We do a sitting and use MRI and use pet, although we don't have a pet, it's scanned on in Gothenburg we do that collaboration with our teams in Gettysburg or in Karolinsk Institute for. Human prayer. And obviously this is quite obvious, but I think sometimes missing that the, of course, if you really just great that you can follow that kind of animal over time and the time points, which is sometimes just nice to have but it very often if you have a disease if you have prophylactic treatment as you start treatment and then you do a disease model, you can always, most of the time just keep them ourselves towards and see what's going on. But if you really want to see on top of disease model you have to introduce disease model that gives you use variability you don't know when and how much there is. And also, it's our highlight that especially if your sector or most of the study will be acute, obviously you have to you don't have always the lecture to follow them to so but there's also very low situation, but imaging gets a unique information that can only be contained in a living animal, like ejection fraction in the heart of a mouse, or renal profusion or things like that that has to be done in a living animal. Because at the end of the day, we also say that the, in a farmer your everyone is stressed, you're all stressed, you know, we don't do any nice to have, and we only though each study has to make a clear decision point, progress to the next stage, and this is, so those are obviously the obvious things that of course as corn said, by distribution kinetics are very important and of course, small molecules you can use pet, we're not, none of those thing. Macro dosing and talk occupancy, percept occupancy, of course, start establishing brain but it can be more and more interesting in organs without such a great. You know, you use. Antibodies, we use stem cells we use cell therapies that you have car key cells that you want to track, so you're very cocky sounds and earning up to a car keys go to the to the tumor or not. And that is quite a few. Yeah, and then somewhere in between, which I like is also high rest and 3d tissue machine. Of course we can do histology tissue machine, but if you want to do all the old structure. You can see bone structure or basketball structure. You need a nose and nose and nose of slices to do someology and never the really good option so the 3D structure I would say, imaging, although those are done. So some of the scanning is done after we have to make the mice but we prepare the mice before and then we scan them. Okay, so. So this is I cross the call it model from my leg molecular to man man is as far as you go. It's all the way from molecular molecules to the to the man. And the first thing I would like to say it's obviously. We always go with tissue machine. So in our team that are very specialized and high in this imaging and must be MSI imaging multiplexing machine so. And it's key for it in for decision to make a proper decision from the legal studies to know exactly what's going on so very often we use the video imaging at a time course in dynamics and then we can mice and then we basically see which cells is our actually targeted or the cells if you run my contrast with college and for college and that's the category actually ends up with a college and so on. So, very often you do an email study, a nice correlation with your disease model, or time course on a group level but it shows something completely different that we're looking for it we don't study now in in cardiac metabolism, for example, a cardiac fibrosis, beautiful correlation to the sea stage, but at the end of the day, it's basically just measuring the arena clearance. And, and then of course, we need to quantify we did numbers, not to. And that is also day over to the laws of AI. What on here to compute. I was a fix better to tissue machine that we were imaging but yeah some examples, when we use AI really to get to improve the. I true the true book of the machine. So, okay, for the first question is obviously by distribution kinetics and bodies. We do a lot of gene therapy. So how do you deliver mRNA to a cell, how do you deliver mRNA to, to the, to the line and then you, you want to compare LMPs for AV. So then obviously we use a lot of I this more as every day I would say that's right. And, of course, a. There are no such things that you see small, we need to see this mechanism model as far as that. But step before that is target engagement at least if you have some target that you have something you know relate to your target something because it might have nothing to do with the disease but at least he knows that you hit the targets and then you can use that to make those different man you can see what those we need to hit the target in the various organs. And of course understand biology and of course we have big transgenic group that make genetic modified mice. So how does this genes effect the function of mice. And also said there are no disease model, though, but there are disease mechanistic model model that mimic some of the ceases of some mechanics of the DCS and of course then I usually it's really key I would say very important when you want to validate this model, because when the motors about a day to you probably run it using other methods, but the first time course of time in the animal, you know what time is the is the really tumors starting up. And my last one basically a farmer has always been in the last year being a vitro cell cell human cell things like that but I take it going back now and we will get more and more realised importance because if you don't have a good mechanism model, it's likely success in clinic is very low, unless you can show something that really elevates you. And of course all people thinking is efficacy studies. We don't do any top studies, I would say, because in top studies, so you need to know, you don't know where what stops finding you will get so you need to cover everything. And it takes a lot of time and has to be very established, but we do a lot of investment is safety studies. When we there is a safety concern, but of both studies one moment with GI effect cardiac effect or renal effect or liver effect. And especially here is that if you find some talks finding, you don't want to just stop approaching immediately you want to mitigate that effect you want to say, yeah if I do some twice a day is it once a day then we can simply lower effect and then in between. And then biomarker next one as well. And of course, as I say, you can of course use pre clinic I think it was more before use pre clinic to cannot set up new tracers of course pet tracers or you see a kind of functional measurement. But at the end of the day, I think most valuable thing is that if you develop a fat pet probe or something like that, you want to see that you want to see a pre clinic model that your treatment specific treatment have an impact on that specific biomarker. In the specifics, I think, and then that gives a leap of faith to get in the clinic. Obviously, because it just tries a long extreme expensive but if you can show that your. Whatever compound your individual have an impact on fibrosis pre clinic and you buy market and you that guides you in a clinical sense. Right. Okay, so a little bit focus kind of center around what we're doing hot fairly CKD respiratory Nash quality self happy and safe just basically the areas we are running in Gothenburg where we're trying to cover all of this area so. So we don't have failure, always use in all studies, also some MRI put them for jet infraction and more likely the more into more roses and the color function of the heart. CKD very much local renal confusion so what happens in after schema in in a kidney, how could you treat those how could you improve the respiratory. CKD rotation was much more than Leo was telling me nothing about a few years ago, but more course to look into people into time for roses. Nash, the models are basically 45 weeks long, so it died in use national so almost a year long study. So of course as you're going to understand it's very critical that you can follow those mice in time. So you had to wait for the week something to get your read out, looking for oncology. Basically, we did in get them I did a lot of transition mice model so there is conventional model all models basically you take a human tumor cells, particularly, and then start growing a big human and have a caliper to measure the the tumor size but now it comes to immuno oncology, which will always be hot area, that's really, really breakthrough. This model, like a lot of small tumors, for example in lungs, how do you capture that to that obviously a machine is a critical self therapy, a lot of focus on self therapy. If you put a cells in the heart, but it doesn't end up. So you can, how do you and how do you keep the cells in the mowing heart of the mouse. We do a lot of that. So I guess, as I said, a lot of phone studies basically study every week. So I think it's just to be some example because as you know in clinic city is the golden standard that's basically everyone around it gives you all the information you need you need to structure information you get basically the function all this kind of so it's obviously what will be around you can. And so but the clinic we can of course we can use my microseed to scan them along, but the resolution is a relative resolution, of course, is for you to see global phenomenon like emphysema or this is the one with the in my sense, three days and this is on the unit. I should try to do my same just one long course we can follow that. And you can see, but as you see, the resolution is much much more and then for example in clinic paramedic paramedic response mapping is the what you want to do you basically do an initial inspiration or expiration, and then you would use the morphos and see how the different each picture change from inspiration and then you can you can look for small ever disease, air trapping, you can quantify the roses, but that requires that the theory you can do that to the MRI because MRI has time resolution but again still MRI that's only one slice to get this resolution we would like to do it in treaty. You know, you need basically a couple of hundred milliseconds per projection so you can do this. This is a different example where time resolution times space resolution is not good sufficient in the city. Hope you will help us with that. And this is just an example of similar to current showing we can use the MRI. Because I said in clinic definitely city but then pretty clinically. I think we go for MRI because it's more visible and you can use it so many decision make. You can use the left because you can see the follow the time course. What happens with the seeds. You know, basically that you go early on and be a mice and all it's just for this just information that seven. So there's no better. There's no process there. You can use this to say they seven. The information that comes there is correlated to the to the roses that will determine mice three weeks later, two weeks later, which means we can use that to to randomize the groups as there's no skewnness in the group you can also, you can select those mice that are non responders. That's to position and then, and of course, if you wait later on day 14, which went on then information goes down and rose goes out, then you can start saying, yeah, we can follow this and if you have longer to start more than three months or something. Then you can measure for that. And that's the quality improves here and. Okay, this is just another decision point, a typical example. You know this mice model, the genetic modify you don't know whether if or if they get tubers and when they get the tumors. So then you use MRI to scan them. This is a routine nowadays. You can see the enumers and starting point we find your sufficient number of tumors, then you start treatment. You can see if you don't treat them and you will produce. Treat, treat that obviously the human disappears and then you stop treatment and see what happens. You can try holiday and then of course, the tumor come back. So there's another question, what is the recipe, even if you treat five patients, you stop treatment will it come back. This is a typical application for given place again, as you can see the type. Okay, so you just want to add this one, I think it's kind of online and so it doesn't really require the mice to survive. But so this typical sample is use. We take the mouse, if you sit with it. If you change all the blood vessel feel all the blood vessel with extra dense costing, and then we take you up to orgasm and we get around this on that this bone healing. This is a new association treatment in diabetic mouse for the news. This is the coronary and mouse of my cutting function, and how could we improve that. So this is another typical sample we do it, used it for kidney disease so we can see, you can now the nice thing is actually this my these images are roughly I would say three micrometer resolution and hold and hold kidney so they are 100 gigabytes I would say from the beginning the big flags, but you're in security are used to that. In this case, you can quantify the different basket or any different compartment of the kidney and then see the effect on models on treatment. So, when do we do you see machine for decision making of course, the first question is I say we never do any nice to have we never do. We always do the easiest way and the machine is very sell on the easiest way to get the information. So we use the machine where there's unique information that cannot be obtained in any other way by ejecting fracking the mouse for example. You can do this prophylaxis measurement if you start treatment and then do your disease, then most of the case you can just kill the mouse and do this knowledge here even into. Now you must not into my chief microscopy or you can do many different. And, and of course, try distribution and drug delivery talk engagement. And, yeah, and sometimes as I said you can do astrology you can do slices but to do those in 3D. So, those are so that even if the if this is just to say, organ delivery of fires to live on mRNA. It takes it and I was it takes you basically five minutes, you scan them out to kill the mice with organs, and you can have really a lot of compounds whereas if you do with strategy, you have to wait three or four weeks and then you get the data one is to Thank you.