as a bit of echo?

 Well, well, Mutaz talk about. Yes.

 Well, well, Mutaz talk about. Yes.

 Thank you so much for giving me this opportunity.

 For me, it feels really, really special to be back in in Madrid.

 I'm really sorry I have to do this in English.

 After ideas in the UK, I've got used to giving talks in English.

 I think if I did it in Spanish it would be a complete disaster.

 And starting from, how do you say, optically pumped magnetometers in Spanish, I have no idea.

 And so, yeah, just do you I'm going to give you a bit of introduction of what Meg magnesium.

 So photography is used for and also the limitations and why we're doing this.

 These, these work.

 Then I'll introduce the, the optics, the optical upon magnetometers, the sensors that we're using, and then I'll go on to talk about the journey of the technology showing some demonstrations and then finishing with the system how it looks now.

 So a bit of an introduction to the neural neuroimaging neuroscience field.

 As you probably know, there are many neurological mental health conditions, so I'm just going to go through a few of these.

 For example, epilepsy.

 Around 60 million people worldwide suffer from this, which is a very debilitating disorder.

 Also, one in 100 people suffers from schizophrenia.

 One person every 95 seconds is admitted to hospital after having had a concussion.

 One in four people will suffer a mental health disorder at some point.

 Also, Alzheimer's.

 In an aging population, dementia is growing markedly and we don't really know what the underlying mechanisms are happening in the brain.

 And the same happens.

 We know relatively little what happens in the brain from the moment we are born to going through childhood and adulthood.

 So sometimes the techniques that we have available to study some of these conditions like MRI or Medigap, we are looking at the structure of the brain, but sometimes the structure looks completely healthy.

 But we need new ways of assessing the function of the brain.

 So looking at the brain activity and one of these techniques is to make magnets is for photography.

 So the name means magnetars.

 So magnetic fields measured from that come from the head.

 So like any other electrical current that produces magnetic field, our brains and the neurons in the brain produce current electrical currents, and we can pick up the magnetic fields that they create.

 So the systems look a bit like this so we can get a map of magnetic fields on the scalp outside the head, and then we can reconstruct these magnetic fields that we have outside the heads, doing some mathematical modeling on source localization.

 We can then get the current that is has produced those magnetic fields so we can get these images of current density and we can overlay it onto a magnetic resonance image to to look at this activity whilst the participants perform a task.

 So some examples of how Meg can help us understand more about the brain.

 So it has really good spatial resolution.

 This is an example of a paper from a few years back where subjects were asked to to move their index index finger on their little finger.

 And we know that these regions in the brain, in the motor cortex, are separated by very, very small distances.

 And Meg can pick up these differences.

 Also clinically, Maggie's were used worldwide for epilepsy.

 It is used to localize the abnormal part of the brain where the epileptic seizures are coming from, and then surgeons can base these there.

 The resection of the part of the brain, they can based it on the on the MEG results.

 So it's been shown in a paper on a thousand patients that the outcome of having how to make a scan before the surgery is it it gives it gives higher chance of seizure freedom in the future also in a mental health field.

 There's also this is an example of some work done in Nottingham a few years back looking at schizophrenia and comparing it.

 A group of patients with schizophrenia compared with healthy subjects.

 So here on the right there, it's a plot that we normally look at in Meg where we have frequency on the vertical axis and time on the, on the horizontal axis.

 And this is a very simple task.

 We just ask people to move their finger for 2 seconds and then stop and we see a decrease in the activity during that movement of the finger and then followed by an increase over baseline, then going back to baseline.

 And if we compare with the patients with schizophrenia, we already can see some qualitatively changes.

 Like if I go back that increase followed by the cessation of movement is decrease.

 So we can already start to pick up differences in some patients groups.

 But Meg has problems and that's why I'm giving this talk, I guess.

 So the magnetic fields that we're measuring, they're really, really small.

 This is sort of a scale of magnetic fields that we have out there.

 So an MRI scanner is a few Tesla.

 We're comparing it to a billion times smaller than the Earth's field, which is what the field we're sitting in now.

 So measuring the max signal is really, really tricky.

 And we need to do this inside a magnetically shielded room.

 So max systems are housed in these rooms to attenuate external interference from the earth, but also other interference like cars and computers and everything.

 You have it around us.

 The hardware, the sensors that are used in Meg are called Squid's superconducting quantum interference devices, and they're really, really sensitive to magnetic fields.

 So you have around three FEM to Tesla.

 So we're in the region of measuring brain fields and we have so conventional systems, the ones that we have there's here in Murray very there are hundreds of these squids and that are placed inside this big dewar that has liquid helium inside because there are superconducting sensors and this is the adult version and this is a baby version as well.

 So squids, they're really sensitive.

 They can pick up the magnetic fields from the brain with high accuracy.

 They are they have high bandwidth, whole head culverts.

 These systems have have been around for four decades now.

 They're well-characterized.

 They're the the occurrence of these systems is that the distance from the sensitive sensor to the to the head is is quite large.

 They are fixed in in location.

 They have to be inside the Dewar and they need cryogenic cooling.

 So this translates into low signal to noise ratio because the fields that we're measuring are really small and we're quite far the fixed sensor location means it's a one size fits all.

 You might have the adult version, the baby version, but it means that the brain to sensor distance is going to be bigger for children, for babies, if using the same system also means that you need to keep still.

 So that's quite challenging for scanning kids or adults with movement disorders.

 And also the cryogenic cooling means there are high running costs associated with it.

 So this brings us to talk about optical power, magnetometer is a few maybe about ten years ago, advances in the quantum technologies.

 So a rapid development in these sensors and they started to compete against squid to measure magnetic fields, very coming of shapes and sizes.

 These are from these fan groups in the U.S.

 But it was really the militarization of the sensors that made the MEG community to get excited for these new type of sensors.

 So these are some examples of the groups around the world developing different versions of of the sensors.

 And also some companies started to do to join the the efforts of making these sensors also commercially available.

 So very briefly their work in principle of an OPM, we have a photodetector a glass cell with some alkali atoms inside a vapor of atoms and a laser.

 So when there's no and when there's no magnetic field of laser, the the atomic spins of these atoms are all randomized.

 Then we shine a laser.

 This is the the pumping, the works of the optical pump magnetometer.

 So we pump the atoms.

 So that means that all the atomic spends align with the laser.

 So at the photodetector, we measure the highest.

 So it goes through the glass cell and we see the highest transmission of the laser.

 What happens when we have a magnetic field, like, for example, the brain near the sensor, the atomic spins are going to change and this is going to translate into a change in the laser light picked up by the photodetector.

 So by looking at the change in light, we can also know what the magnetic field caused This change.

 So comparing with hopes, the sensitivity really depends on the size of OPM's on their attack showed many different groups developing them.

 But he's getting very competitive, very similar to squids.

 The bandwidth is a bit more limited, but enough for measuring magnetic pressure on the brain.

 The number of channels is increasing quite rapidly and multichannel systems are being characterized.

 The good thing is that these sensors reduce the distance to the scalp.

 They are flexible so we can put them anywhere we are not limited by the viewer, They're not kept inside any anything and we don't depend on liquid helium.

 So this brings me to talk about the the journey of the work that we did during my Ph.D.

 in in Nottingham.

 And the vision back then was can we turn these big, heavy, ugly machine expensive into something that we can wear?

 We can put sensors on a helmet more similar, like an EEG, like a electro and photography device.

 So we started doing some simulations which basically men can we bring these?

 If we bring the sensors closer to the scalp, we're going to get higher signal.

 We're going to wear closer.

 So we were going to pick up more signal.

 And that overall meant that we get a theoretical advantage of a55 times more signal just by putting the sensors sensors closer to the head.

 It was later in 2017 when we did our first one channel recording with a sensor four from Houston.

 And these think Brian knows this is Gareth Barnes from UCL and we made this helmet for for him based on, on an MRI.

 So we knew where the sensors where in respect to so his brain and we proved experimentally that it was about four times higher.

 So agreeing with the simulations in blue, you see the LPM signal and in in red you see the squid signal multiply by four is about four times higher for the OPM.

 And then he was later adding some hardware to the to the system that we did our first wearable OPM recordings.

 The problems is that so as you saw in this photo here, we had the subject clamped on to the bed.

 So it was not very comfortable and it really defeats the point of flexibility of OPM.

 So we wanted to have people with the sensors free to move their head.

 But if we actually if we move the sensors like is shown in this video, I just go and take a drink the the signal of the OPM's satellites and we we don't get we don't get we cannot use that data because we've exceeded the dynamic range of the sensors, which is quite limited.

 So we need to provide a better magnetic field environment within our room.

 So this is when my colleague now home joined and he he designed these spy plane coils that you can see these photo there originally that were made to fit around an existing cryogenic mech.

 But Niall designed these by applying coils to to create ascension here.

 So to create the magnetic fields needed to compensate for the remnant field in the room.

 Actually, these designs are adapted from MRI methods.

 And actually so Peter Mansfield's work in the building that we take his name in Nottingham is based on some of the developments, some of the legacy of these these work for MRI.

 So thanks to those coils, we can reduce the magnetic field inside the room to.

 So it was around 30 so monitors and then we go to pick a Tesla.

 So much, much smaller.

 This allows the sensors to keep working even when the subject moves.

 So here you can see a video of me putting a ping pong ball on a on a bat.

 And we can see that the head is completely unrestricted where there's a bit of movement.

 But still we can see we can get really high fidelity data even though we have the head moving.

 So reconstructing to the motor area of the brain and then batting the ball for 10 seconds, we see that decreasing amplitude followed by an increase.

 As I show to in the the other light.

 So this was all very exciting.

 And after that we started exploring all the ways of of doing magic.

 So before scanning children below the age of eight was really tricky because they don't, they don't keep still.

 So we did our first pediatric experiment without the two brothers, a two year old and a five year old with a modified helmet that we bought on Amazon and just put some sensors.

 And we did.

 We got really good results.

 This was just a sensory task with a mum, the mother stroke the hand of the of the kid and we see that changing in, in activity in that sensory part.

 And then we wanted to play a bit with virtual reality because now we, we allow people to move their head than can we immerse them into a Yeah, a virtual virtual reality environment.

 This was really tricky, but nevertheless we managed to get some good data using this headset and then luckily the sensors became much smaller.

 So this is that.

 So the second generation of sensors are just as the size of a of a Lego level piece.

 So we started in 2019, the first experiments with this second generation.

 One of the first things we did was to compare it with EEG because these might be of interest in some clinical work, perhaps.

 So we did.

 We had an EEG cap and then placed to ops on top and perform the same the same task we we put the sensors on the the optimal place to get the, the, the motor response that we were going to do.

 And we got really similar data across EEG and on OPM.

 However, we saw that when moving.

 So if you have a subject moving the EEGs much more susceptible to to muscle artifacts, also the cup might be moving and the the data that we have them, the sun electrodes moving on the head, that also degrades the data.

 So we saw that he was about ten times worse than when using the ops and then that by the end of so towards the end of the year, we had a 50 channel 50 sensor system.

 So I'll I'll show you some of the demonstration with this multichannel, multichannel system.

 So this is an old photo now we moved, so we got a bit of funding, so we moved to a new room which is dedicated to open mics.

 So the room that we were in before where we had a a CTF system, one of the cryogenic companies there, the field there was 30 Nano Tesla, so quite large for the ops to, to operate these rooms that are manufactured by magnetic shields.

 A company in in England they they started working on making these rooms with lower magnetic background field.

 So we're able to to to use the census which we start with the census a much better place.

 We still have these.

 So you can see here a newer version of the biplane A coils that you can see in that photo there on the back.

 We also have gone through a few iterations of helmets, so we started with these subject specific ones and based on MRIs from the participants, then we went through a bike helmet, something that kids could wear and flexible caps and then more 3D printed, but now generic that more people can use.

 And later on I will talk more about this.

 But with Serco, we've been making these generic helmets that are lightweight, so we them multichannel systems.

 So we have 50 now sensors.

 We we did comparison one with the cryogenic, which has about 300 sensors, but nevertheless we found that.

 So this is we scanned the subjects six times in each of these.

 So we've created helmet with a flexible helmet and then in the cryogenic and during a very simple task, a visual multitask, we see the visual response and the motor response.

 And very so these are all the different runs, very repeatable results.

 And also in terms of of in our comparing 50 sensors ops with 300 cryogenic sensors and then also use this data to look at conductivity.

 So it's so quiet, I would say hot topic in Meg to, to look at connections across the whole brain and in the different frequency bands and different regions.

 So we got very similar results comparing the OPM to the Cryogenic as well.

 And also it was here, it's probably easy to see we could pick up also subject differences.

 So two subjects were scanned multiple times and this is the connectivity that the OPM system is picking up for both subjects.

 And this is the cryogenic and both OPM and cryogenic pick up differences in the same region of the brain for both subjects.

 Then we've been also doing all the types of experiments.

 So Motorola and and this is something that my colleague Ryan Heil is working on.

 So paradigms that we cannot do in a conventional.

 Meg, if I go back, you can play that video that requires a very large head movement to be able to to look at the instrument and play it, and we can get this once again.

 He scanned a few subjects multiple times, I think every day for a week, and we get really, really good data repeatable results, even though we are making very large movements.

 Then something more exciting now on this is quite new.

 Nigel Holmes.

 He's been developing new types of of coil systems.

 So the ones that you showed a biplane across which can be seen here at the back that I'll show you.

 I think there's another photo later on but he's been working on different types of coils that can tune can bring the field down following a subject as they move.

 So in here he's moving randomly, sitting down and turning around whilst he's pressing buttons and still he can get decent data even though he's moving around the room.

 But you can pick up these changes in the brain.

 And again, we had to do it this time.

 Not just me, but playing against my boss.

 Matt We played ping pong the two of us, and here you can see these coils that I was talking about.

 So these new coils can noodle the field in two different regions.

 So where the two heads are allowing for these movements of the head and we get really good also.

 And data in the motor region whilst both both participants are playing ping pong, this is some work using the same sensors that we use from Q spin.

 This is from Brussels and from the group of Xavier where they've recently published a paper where they actually measured epileptic seizures with the OPS, which is really exciting and really exciting finding that it hadn't been done previously.

 And so this is another application of ops and also the flexibility of opens gives us not only ways of measuring the brain activity, but now we can start to think of measuring the the world, the human heart or the, the, the fetus heart.

 Hopefully the brain, the fetus brain as well.

 So we've done preliminarily some experiments where we've picked up the the maternal heart on the fetal heart with open arms by placing them on the on the belly.

 But yep, this shows that opens gives you also different different ways of measuring by magnetism from the from the body.

 I mean this is a bit of a new so we I show you the first generation of sensors that they were quite large and we then got the second generation.

 This will be the third generation of sensors.

 So now before we had what we call dual axes because we can measure the feeling to direction.

 So we place it plays the sensors on the head, so we get a radio and a tangential component of the field.

 Now we have try actual sensors.

 So there's another.

 So the laser that's in the sensors is split and you can you can pick up and you can measure the the other component of the magnetic field.

 And this in a paper by a few years, couple of years ago showed that these can help with rejecting interference better because you have more information of what the fields look like also gives more uniform coverage so more homogeneous sensitivity, especially in case I'll show you in the slide in a moment, and also gives more signal because you have, again, more information of the fields.

 So this is the the theoretical advantages of traversal.

 So in the middle here is the radial sensors.

 This is what all the experiments that I have shown you with the previous sensors.

 So this is for an adult and then for a two, four year old and a two year old.

 And you can see that the pattern, the sensitivity profile starts to get quite patchy, but we use its actual array of sensors.

 The it gets more homogeneous, so we're picking up more signal from the brain.

 And we've done a few experiments of these Dirac shows just to check that we're still getting a really good data and this is some work from last year we can actually we can pick up.

 These are a single trial responses, so really good sensitivity from these fractures.

 And we can now get a 3D visualization of the field that we couldn't do before because we didn't have that extra component.

 So we the tracks are open now that we have in Nottingham.

 We've been doing more new tasks that require large head movements.

 So this is the work that Mollie Wray has done with a 90 channel array of track shows where subjects were asked to, to to write down a word that appeared on the screen and that involved large heft movements.

 But again, the data was looking really, really good.

 And something that I particularly I'm really excited about and is the is the pediatric make that we can now scan children that we were not able to do in conventional systems.

 So this is work when Ryan, Natalie and Lucas are involved and so we've been developing these later helmets with Searhc and other companies and there's an ongoing study where we are trying to recruit children from, well, newborn to 13 years old.

 We've scanned 11 kits so far and we've done two tasks.

 So one of them is this is a paradigm that is used in sick kids where they they scan children and typically develop children, but also children with autism.

 So one of the tasks is just a few faces up on the screen and also some concentric circles to look at some gamma bands or higher frequency responses.

 And then there's also we've done a sensory task where these Braille stimulators stimulate the index and the little finger, and this results are looking really promising.

 These are the emotional faces.

 We can get large evoked responses and at the back of the brain, in the visual and primary visual areas and also with high SNR in the visual areas and also the FUSIFORM and the Braille paradigm as well, we get a and evoke response.

 And also, as I've been showing for a while now, they this drop in the frequency band during movement, while during the stimulation story and B to band or localize to the sensory cortex as well.

 So now I'll go through what the system is now and going through a bit more.

 My involvement with Circa Magnetic. So and clearly there was a lot of interest in OPM technology and we we wanted to, to set this up to and to facilitate this technology, to go to other places, to other labs where they can do their their their specific research topic.

 So we with the University of Nottingham on the company magnetic shields that make the the enclosure we formed circa in July 2020.

 And so far we've installed our system with so 44 dual axis system in SickKids in Toronto, a 64 sensor in young epilepsy in the south of England and very recently in Boystown in Nebraska, another 64 system.

 So I'll just briefly go through that.

 These three.

 So in SickKids, the Hospital for Sick Children and we retrofitted we call it a retrofit system because they already had a shielded room there from a from an old cryogenic system.

 You can see that at the back.

 So we installed the biplane, our coils to go around and to bring the to to compensate for these changes in in magnetic field and their main study, their main research is on autism.

 It was probably not the easiest site that we could have started with after forming Saga because like here, the CTV is in quite a quiet, I'm getting sicker and in a very quiet part of my life.

 But here in SickKids, it's in our city center hospital where there is lots of interference from elevators to the metro to building works.

 I think the mighty of garage.

 So basically everything you don't want next to a mic system.

 So it was quite tricky.

 You can see these blue lines that's just the sensors in the room not doing anything.

 They go out of their dynamic range very quickly and compared to Nottingham with where we are in in the university campus in the middle of green fields, there's nothing around you.

 You don't see that much interference, but with the coils compensating dynamically for these changes in field, we can bring the, the sensors down to their operational range and we can perform, make experiments and which would be interesting the way.

 Yeah, as you say.

 And so yeah this is so Ryan went out there to set up this system in Toronto and he so he scanned himself five or six times in, in Nottingham and then he said kids and we, we see sort of similar from that paper from a couple of years ago, very comparable results across to two very different sites.

 One very quiet, magnetically quiet and one, yeah, very, very different.

 The other sites in young epilepsy in, in the UK, again this was a retrofit so there was already the the shielded room there because of a project from well between magnetic shields Nottingham UCL and young epilepsy.

 They with this a UK project of light new room.

 They wanted to make a cheaper, lighter shielded room.

 So getting rid of one layer of the the the metal that they're made of but having coils you can see in this photo here having the coils inside the walls of the room.

 Can we give enough back around magnetic field for the sensors to operate?

 So then we installed a 128 tunnel system there and the then the patient supported the helmets and the the software and what they what they're going to be doing various they will be doing some clinical trials involving scanning children with epilepsy.

 This is some work that's been going on there as well.

 So doing concurrently, OPM's on an EEG.

 So this is work by silica where you should use the 64 sensor helmet and then some electrodes on the scalp and you get can very yes, similar results from it make an EEG in a simple task of just eyes open, eyes closed.

 And recently I have two of my colleagues are currently there in Boystown, in Nebraska, finishing the installation of this or the third system, which has been the first one where we've done all of it.

 So the installation of the room now you can see there, which is the same as the young epilepsy and 128 tunnel system.

 And they recently they've done very nice videos of the the inauguration of their of the center.

 So I think that's the I'm nearing the the end so to conclude these new generation of of sensors of quantum magnetic field sensors have enabled these wearable make magnets and so photography to become a reality.

 And now we can have participants moving and doing some different tasks that they were not they were not able to do before whilst getting really good and high fidelity data.

 And also this technology opens up new ways of doing new neuroscience research, but also hopefully help in the clinical setting as well.

 And so finally, I think for me, bringing open that to Spain is like closing my circle.

 I started, I studied here, I went away and now for me it would be a great it will be a dream to bring opium to Spain, but not appearing on a trailer offer of Amazon Prime series up here in Spain.

 Completely random. But it was.

 Yeah, it was quite funny to see.

 But for me, actually bringing that technology and putting the system in the industry to be would be, yeah, a really a pleasurable dream.

 A dream come true.

 So with that I'll just I want to thank all of my collaborators and colleagues and funders and to all of you for listening