We've got Hagay next. He's extraordinary. And I'm absolutely thrilled to have him here. I'm gonna do the intro because he deserves it. And so he and I have been speaking for about two years now when we first sort of first got to know each other. So Illyria Health is a it's venture capital backed. It's a startup in Israel. It's making it is making waves globally. I know I know some of the international stuff will learn more. It was inspired by his own personal experience with ADHD and navigating with his children. And at the moment, they are validating software around continuous passive monitoring around a smart wearable. And actually, of the things that I thought was really interesting about Margaret's.

Part was where she showed the sleep diary. She was pretty old school. And we didn't have time to ask. I wanted to. What do you think of a wearable? I will give that information direct over. Right. K. Without further ado, I'm bringing him in. Hello. Welcome.

Pleasure to be here.

Well, it's wonderful to have you here. And it was great to see you in London a little while back and talked over video many, many times. Thank you for being part of the conference. And when 2023 schedule started coming up, I was so enthusiastic to get you involved because you're also doing primary research through the use of wearables. It's properly globally revolutionary, it's right at the cutting edge of technology. It's very, very interesting. And I was always the very first time we met and you showed me some of your data, was wanting to get that to a wider audience. Do you want to do a quick you've got slides.

Yes, I do. Yes, I do. Yeah.

And similar to to Margaret, we've got you've got me delivering them for you, which is yeah. But do wanna do a quick introduction first? Talk talk about you, and I'll get the slides up. Okay? Super. So Hagay Levy, as you mentioned, from Israel.

Most of my background is actually not from the health care side. I'm a finance guy. I worked for many, many years in the capital markets, many years at the high-tech industry in, like, multiple positions. I did product. I was CFO for many, many years in multiple industries. But I think as you mentioned, aside from that, I'm also a father. I have ADHD myself, but I am 54. So, you know, in my generation, who knew? But aside from that, I'm a father. And together with my wife, who is actually my cofounder in. We have four kids. And as of this morning, all four of them are diagnosed with ADHD, slaciated. Two of them also diagnosed with anxiety and depression. And we noticed that as gatekeepers, there are many missing links in the ecosystem.

And, potentially, the the the main one that actually we were bothered with is the pharma side of things because, statistically, the pharma is the best line of defense. Together with that, we had zero data with regards to the efficacy, safety. Same goes for the physicians because they're basically relying on our continuous data flow, object subjective one. So this is me, and, actually, that was motivation to even before establishing to initially start searching for a solution that will allow us to get objective data while analyzing the titrations, calibration processes.

Okay. I'm pleased that your so slides are are ready. And and just to let people know, you know you know, the slide overlay still has the donation overlay. So that means you are absolutely able to, should you so desire, to to donate. If you donate 20 pounds, you'll get confetti and 50 pounds, it's really just a celebration of you. So 50 pounds gets fireworks. The yeah. So do donate. That'd be lovely to see. Obviously, we get no money from government. Everything we do is is funded from donations, so very gratefully received. So you know I've got to do it. Two things that struck me just in your little introduction. So slides are ready. We'll move on to them very fast. One of the things when you first talked to me, you said that you had a.

Conversation with a child. It was around a child. They were diagnosed with ADHD, received medication. Is hopefully I get this story right because I repeat it a lot. And the teacher was basically, like, it's really, really helping. Like, it's super, super good because the traits around hyperactivity was significantly altered, but the kid wasn't learning. And there wasn't the change in in that. And what you recognize was the inattentive aspect wasn't being looked at. And a change of medication suddenly saw the child moving through year reading ages. But the but and but that child could have languished. Right? It it the teaching teacher was happy. It could have stopped there, so it wouldn't have been getting.

The actual help. And I think that was significant for you and when you realized that trying to learn more about how the medication's impacting the child and doing it in a more sophisticated way was critical.

Yeah. You just Hopefully, that wasn't slides two, three, and four. No. No. But you I mean, there are so many problem in trying to navigate in this, like, of of trying to understand what's happening off meds, on meds. You mentioned, like, a personal case, which is I call it, like, a false negative alert by the teacher because as you mentioned, it was my child in the when he when he was in the first grade. And, I think you can see some measurements in slide number four. I mean, this one. And for her, things were okay because the activity was kind of low. The impulsivity wasn't there. He wasn't bothering anyone, but he could not read.

After, like, the middle of the first grade, they just she'll see she said, again, I cannot blame her. She's she's a teacher with, like, a huge class. She said the medication is working. The kid is not hyperactive, but he cannot read, so let's move on to special education. Let's bring assistive teaching. And the only problem was that the impact of that specific medication on him wasn't that high. Practically, when we calibrated, the kid started reading in a day. And, actually, currently, whole life story is different because he's now 13 and he studies at university biotechnology. And again, the only thing we're just putting him on track in the sense of finding the right medications.

What happened back then was that we felt something is off in the analysis that she brought us, and we started measuring. Actually, I have it here with this Chinese wearable with only two sensors back then. And the data guy, which I you're gonna see in a second, told told us, look. I am not a clinician. I cannot tell you whether the medication is working or not. But when I'm looking at the physiology and the stimulants should change your physiology, I don't see any change. When I'm trying to aggregate all of the data and to separate the situation between the off meds and on meds, I cannot separate the data. For me, it's the same child. And when we tried a different medication.

It told us upfront. I don't know what will be the impact the academic impact by the kid. For me, it's not the same child because there's a huge gap between the off meds and between the new medication. So, again, that was like I call it like a false positive the the the so response negative. So many other problems. Let me just link on on this child two years afterwards. The teacher called me and told me, you should go to the physician tomorrow morning. The medication is not working. The kid is killing me. He has been eradicating the class for the past two weeks. Do something. So I measured him.

And we saw that the separations were very, very strong. So I told her, let's look for an alternative solution. Let's move him to another place in the class. She moved him to another place in the class. This is it. Problem solved. Someone was annoying him. She didn't knew that. And and instead of going to the physician, starting estimation, trial and error process, We measured. We saw what's going on with the medication, and we actually searched for the right solution, which actually was a behavioral one rather than a pharma related treatment. So, yeah, again, there are so many cases. The the engagement problem but but, again, let me stop here and maybe yeah. So if you can move to the next slide. So as you mentioned as I mentioned.

We are in Illinois, we're super focused on practically the pharma side of things. We are kicking off the activity with actually pediatric ADHD even though we are working started to work on other indications, anxiety, depression. But, again, as of this morning, the all of the activities focused on ADHD. Next slide, please. Yes. So as I think I mentioned, we don't aim to interfere at all the initial diagnosis. We don't want to be there. Same goes for the therapeutic side. We are definitely not the we are not curing the ADHD. I don't know if anyone can cure ADHD. But we are if you can click, we are trying to if if you can click on next, We are trying to shed some light in the never ending calibrations.

The trations process that again, it's a never ending process. We just started measuring young adults in the university. But, again, circling back to the clinical trials that we did and the endless discussions with physicians, we decided to focus initially on the initial several years in which the kid was diagnosed. And now you start the calibration titration process. And, again, everything is manual, subjective. The level of care can be enhanced, many, many labor associated cost. But if if I may choose as a father, actually, if I may cherry pick one problem, I would actually cherry pick the engagement.

Because the engagement in ADHD is super, super low. Again, from the physician's point of view, it's the difficulty to shorten the time to the effective treatment. From my point of view as a father, it's the inability to balance the equation or flex minimize side effects, and you might have side effects because it's stimulant. And together with that, while minimizing the side effects, measuring a proxy at least for the efficacy, safety as well, So we decided to focus on that. Next slide, please. And, obviously, the goal is to illuminate the ecosystem. And next slide, please. The basic notion is the connected patient notion, meaning utilize the new tiny hospitals that we currently have on our hands with many, many, many sensors.

The connected patient notion has been actually up and running in other indications for quite some time, you know, with diabetes, third age indications. You can find connected patient in many other elements. We are adopting it for mental health and initially for ADHD. And, actually, the basic idea that my wife came up with, like, five years ago is pretty straightforward because if you look at the existing science, I'm I'm not even talking about our clinical trials. I'm talking about the dozens of clinical trials that were conducted worldwide, you will find a very, strong correlation between effective stimulant or nonstimulant treatment and changes in physiological markers.

Meaning less movement, less fidgeting, slight increase in heart rate, changes in heart rate variability, changes in the spirit, and so on and so forth. We simply took it one step higher and sense of let's gather ensemble of markers from any smart wearable device. We are not dealing with hardware, meaning any Fitbit, Apple Watch, Samsung wearable can do the job. And, technically, if I may simplify it, we trained algorithms to and, again, I'm super simplifying it, but, technically, we're analyzing the patient mathematically off meds and on meds to, again, share some important objective data points to the ecosystem. If you can move to the next slide, please.

Actually, now we have the initial version of the solution. We we are not in the market yet. We don't want to be in the market too fast. We are actually starting to work with organizations. We are not doing any direct to consumer. We decided to avoid that. So currently, the version that you actually see here with my little son's data is being used in pilots globally. Technically, are using Samsung Galaxy wearables, but, again, the notions to be agnostic. Here, we have movement, angular movement, upgrade, edge of things, some other sensors. And if I may simplify it, the current solution, which has been piloted as two main layers. If you see in the middle, the smartphone, a lot of data points that, honestly, parents asked us to.

Provide them. I mean, starting the clinical trials, we did a lot of sort of AB testing with parents, again, as gatekeepers, and they asked us to see the daily movement in the class as a proxy for the behavioral discussion in the evening, the medication reminders, which is, again, a huge problem as I think some of us are aware. But aside from that, if you can look at the right hand side, we have an aggregated report. It will be delivered on a monthly basis providing a periodic ping. The things are okay or not okay, meaning we can actually separate the data or not. And this report has three main layers. Again, raw data, but now on an aggregated level. For example, I talked about the cardio.

So again, it's not at the clinic measurement with an ECG device, but as long as we have the data, let's expose it. Because I have seen some cases in which the heart rate went crazy, and I told the mothers and the fathers, please go and check that because I do see a huge increase in the heart rate. And then two layers of calculated data. You mentioned my son's case in the first grade, so this is what you see actually in the middle. It's I think it's too small, but you can see on the left hand side the nomads in the middle tending to the left, the Concerta, which as you can see, the separation was kind of lame. And on the right hand side, you see the Ritalin LA twenty, which is actually the medication is has been using since.

So, again, it's it's not only for to avoid false negative alerts and or false positive alerts. Let let me let me just circle to another problem, maybe the most common problem. Again, engagement. The medication medication is doing nothing for me. I mean, let's reduce the dosage to eliminate side effect. And now I don't have any side effects, but I will tell you that the medication is doing nothing for me, and game over. I'm out. So just, again, exposing the data and showing that maybe the separation is not as strong, but still, there is an impact is actually crucial in the sense of trying to convince, in most cases, the adolescents that there is something going on under the medication.

And then last but not least, once the medical treatment is set, we are producing the chart that you see on the bottom, which is the average intraday pattern or impact, pattern of the impact of the medication on your physiology, again, as a proxy for efficacy. Some physic physicians like to like to refer to it as a proxy pharmacokinetic curves. Again, it's not as if we're taking blood measurements and actually measuring the level of drug in in in your blood cycle, but we are using the noninvasive markers to portray the average change in your physiology, again, while you are sitting, doing the cognitive assignment to.

Not only analyze whether the medication is effective when it is at peak, but also measuring the duration, the proxy TMAC, proxy CMAC, and some other elements. Next slide, please. Yeah. Actually, before the next slide, obviously, we are moving along with the development. And as it seems, the next layer of development will actually, I think, do a drill down into the sensors themselves. It's it's pretty hard to see whether the kid is moving more or less. Again, it's very subjective. Same goes for the cardio. It's not easy to actually identify whether it's up or not. But when we started doing the drill down into the raw data itself, we did find some very interesting, as you can see on the left hand side, movement movement patterns.

Again, it's very hard to actually detect whether the kid is moving more or less, but usually on medication, the the movement is a bit more smooth as you can see in the right hand side. And on the on the right hand side, you see some sort of a cardio analysis. Again, I'm just a father. I'm not a physician. I'm definitely not a cardiologist. But we did allocated some pretty, I would say, generic patterns between OFF meds and ON meds in which, for example, the HRV is a bit more stable under the medication. It's not lower. It's more stable. So, again, a lot of data points that we will provide to the physicians in the future. This is for the next version. Next slide, please. So, again, multiple stakeholders.

All of us are in the same boat in the sense of suffering from, you know, from the mess, from the trial and error, from the guesstimation. Obviously, I have the users, the parents, and the patients in the middle because, actually, that was motivation, as I told you, for establishing the company. Obviously, for us, the better caregiving is the number one mission and motivation. Obviously, looking at the payers, providers, you know, HMOs. Obviously, it's better caregiving, but as together with that, the cost reduction element is crucial because statistically, at least in The US, I think 20% of the ongoing visit to the clinics are false positive alerts by teachers or parents. And I think that the engagement problem.

Is important for, actually, all of us, including the nonmedical version of the special education, which we are starting to piloting. Obviously, for the pharma players, again, for obvious business reasons, but, again, even as a as a parent, I'm not an advocate of pharma, but as a parent, I think that's avoiding the treatment. It's not solving the problem. It's, you know, sticking your head in the sun. So yeah. Next slide, please. We've done actually, we've done three clinical trials up to now. I can elaborate on them afterwards if there will be any question. Mean, I mean, just guide me through that, Henry. But roughly speaking, in Israel, we did the initial clinical trial with the largest provider in the country. It's actually a huge HMO.

And if I may simplify it, we then will show that we can mimic the initial baby step that a physician is taking while analyzing the patient. Again, after the diagnosis, meaning we have above six symptoms of meds, which is bad and needs to be changed. And then using the initial medication, it's below six symptoms. So maybe it's not optimal, but it's sufficient. So we we actually train the initial version of the solution to mimic that. And in The US, we're working with Cincinnati Children's Hospital, well, a pretty advanced provider on the looking even at the technological side, and they told us at the get go. We're working there with professor Epstein who is pretty well known.

KOL in ADHD, and he told us for me, meaning for him, there is one holy grail. And the holy grail are these proxy personal intraday patterns, proxy personal PK curves, and other is guidance. We'll actually develop that ability. Next slide, please. Obviously, we are gathering data. I mean, this is not for tomorrow morning. But clinically, there is now a lot of discussion looking at precision medicine in the sense of predictions, in the sense of patient clustering. Because, again, I'm saying stating it over and over again. I'm not a physician, but if you look even you will account for a diagnosis of the patient.

Usually, the main attribute that the physicians will look at is body weight. And, obviously, there are some other attributes that might impact the effectiveness or the the guidance for the medical treatment, the ethical origin, gender, comorbid conditions, obviously. So we are starting to gather data looking at the long run pattern of of trying to aggregate data and then cluster the patients so that even prior to the initial diagnosis, we we we will initially place the patient in the cluster, which is a bit more enhanced than this is your body weight. And by that, maybe impact even the preliminary process of diagnosis. But, again, this is not for tomorrow morning. Currently, we'll state in precision medicine.

Once you're coming with a specific medication, meaning once we see you on medication, we will tell you. We can separate the data. We cannot separate the data. These are the patterns. The patterns are lame. The patterns are significant or not. Next slide, please. So, again, as I mentioned, we are not here to diagnose. It's there are many solutions. I know I know that in The UK, you have where is it? I lost it. I I I think it's hidden by the the barcode. But but there are, like, many solutions at the clinic. We're actually assisting or supporting the initial diagnosis. We decided not to be there. As I mentioned, looking at the right hand side, we are definitely not curing ADHD. We are not enhancing the medical treatment. We are not there.

But we are trying to fill what we see as a gap in the post diagnosis and then never ending calibration process, situation processes, again, getting objective data. Obviously, there are some processes going on now because the monitoring is there, but in most cases, it's manual processes, again, using the cushioners. So this is actually the gap that we're trying to fill. Next slide, please. Yes. So, basically, this is who we are and where we are at this stage. As I mentioned, we are not here to sell anything. We are not in production. Again, we decided to avoid the temptation of quickly going direct to consumer. And believe me, it's it's a big temptation while managing the company.

So we did, like, a preset round that you mentioned, and now we have the initial version of the solution. We did the pilots that you see on the upper right hand side, and now we're doing the pilots which you see below with providers. Actually, the main one is in The UK as we talked about, Henry, with ADHD in the Northern Eastern part of the country. And as it seems, we will initially try and commercialize a solution in The UK as the initial market. I mentioned the special education, and, again, the special education is actually pretty intriguing. I think it's a nonmedical device for many reasons, which again, I will avoid now because I know that we have limited time, but.

We are starting the pilot with one network of special education. They're called the Dillard Education Trust. Again, they are dealing a lot with ADHD as gatekeepers maybe, not as physicians, but still, they are very much engaged with analyzing that. So we have secured the IP and everything. And I I think I mentioned at the beginning, at this stage, as you as you saw, we are super focused on actually pediatric ADHD even though we are we are not saying no to adolescents. But we are starting to measure young adults. They're actually college and university students.

I'm not certain that we actually purify the problem over there. I mean, again, as I told you regarding the younger population, if I would have to cherry pick one problem, at least for me, I I would choose the engagement. We were approached actually four times by universities telling us there is a problem with the students, in some cases, over sedating themselves with the stimulants. And please assist us in measuring that because it's not healthy, and there might be, like, a dropout from university and so on. So the fourth time that we were reached out to it was from the Israeli Tel Aviv University.

So we said, okay. The hell with it. It's it's close, so let's let's do measurement. But, again, this is or looking at the future, we are not there even in cementing the value proposition, to be honest. Next slide, please. Yeah. This is us. Obviously, we have some developers. I mentioned my wife that actually guided me actually five years ago in okay. This is the problem which you solve. So you see here, we're pretty strong on the data sides, but aside from that, obviously, we're getting a lot of assistance from clinicians. Next slide, please.

Yes. So aside from that, I have some description of the clinical trials. I mean, Henry just guide me through that because I cannot see the questions. So I can share some data on them or maybe I'll relate to some questions. I mean, I'm bringing the mic to you now again. You are muted.

There was a lot of accusations previously that the sandwiches were being eaten in the in the background, and it it was not. And they were not. Well, it's actually technically a lie because I did eat a sandwich later, but that was not the reason for the noise. The Yeah. We've had some questions. I think the the bit that really blew me away, and I've seen some comments come in. So we had someone just saying, wow. This is this is my this is my son and pop it up. He couldn't read by second grade, and and the school just said he wasn't ready for school. Like, he just pushed away. The yeah. The thing that blew me away is with the data. So it'd be great. Yeah. If you'd be happy to share some of the data so people can see.

Because of and obviously, it's mostly for children, but it's obviously, it's the same for adults and just that understanding of how the medication is interacting and be wonderful.

I have to say, I can't see I don't know which side is next. I I'm just I'm just I just press next. Share it again, and I will, like, walk you through them. Again, I will do it very, very briefly, not not to bore the audience. But I will just give you give you not only the description, but actually the the basic notion behind the technology and obviously the clinical trials. Because, again, the claim to frame and and the mission that we had was wasn't that simple because the the the the goal was to see whether we can train algorithms to analyze noninvasive markers. And, again, these are just a sample. The goal is to use any ensemble of markers because, again, we are using different wearables.

So, again, the goal was to see whether we can take noninvasive markers and then train algorithms to, at the end of the day, mimic the gold standard, mimic the DSM five compliant analysis of the medication is working or not working.

So so that's So that's a big it's a big deal. Right? Like, essentially, what you're doing is prior to your involvement, people were, you know, doing hospital trials, someone either being hooked up to a big machine or someone coming over and taking measurements all the time, and that's how they were plotting it. That was the clinical trials at the time. And you've gone along and said, right. We've got and you just wonderfully sort of showed it up. We've got that on our wrist, a huge amount of it on our wrist right now. And we can therefore work out not just the general curve, we can work out someone's personal curve. And we can realize how because otherwise what happens, we do this really.

Know, like people did as you talked about and the big challenge with children that they can't answer, that people go like, how's it going? How do you feel on it? And that's actually quite complicated because it's mixed up with everything else that's going on in your life. There's so much. And so trying to give a tool to help express that It's clearly still you wanna ask people, but it was a big deal. Now this first slide you've you've put out, which I now put my ugly lug in front of, is where you were looking at the different things to measure.

So, technically, we train the algorithms to use the basic ensemble of sensors because, again, there are different wearables with different ensemble of sensors. Typically, in a valid wearable, you know, the Apple watches, Fitbits, Samsung wearables, you will have movement, joggle, you know, angular movement, which is actually very interesting in ADHD, you know, the wrist restlessness, heart rate, you know, cardio and heart rate variability. So this is, the basic ensemble that we're using, and then any additional molecule is obviously can only help, but but it's not mandatory. If you can move to the next slide. So, again, we did a lot of preliminary analysis before starting clinical trials before associating.

The separation to the gold standard. We think initially, we wanted to see whether we can actually separate the data. So I keep on circling to the example that you gave of my young son and my wife's ordering me to solve it. So you see, like, on the left hand side, so you see the inability to separate the Concerta and the off meds data with a few sensors. And on the middle, you see lame separation, but still a better separation with more sensors. And on the the right hand side, this is actually, I used this graph because this is my older son. When he turned 16, he suffers from side effects terribly. You know, he's a musician, and he cannot play. And.

And he told me I'm off meds. I'm having side effects. This is it. I'm doing the exams off meds. And I told him, okay. Let's cut the pill to let's use half a dose and see how it goes. So he told me, I don't have any side effects, but the medication do nothing nothing for me. I'm out. So I actually use this measurement. I made him sit and do an assignment, and we measured him and actually used this slide to convince him in maintaining at least the half a treatment throughout his high school years by showing him, okay. It's doing something. But, again, these are not formal clinical trials. If you can move to the next slide. In the initial clinical trial that we did, I mentioned it was in Israel Israel with the Claritel Cell Services.

And if we recall, the goal was to see whether we can mimic the initial baby step of analyzing the patients on some sort of a binary nature. Above six symptoms, below six symptoms, we should the situation should be changed, and now it's sufficient, and maybe we can start the calibration. So we took a group of actually 45 patients, and we had them going through actually several diagnosis processes because we were intimidated by.

The possibility of training the algorithms on almost positive alert. So we used several diagnosis, and then if there was a conflict, we actually ruled out these patients. So practically, we used 37 patients, and we got the raw data for each and every patient. When the patient was sitting at the clinic doing a cognitive assignment, twenty minutes off meds and twenty minutes on meds. And then we started playing with the data. If you can move to the next slide. Before associating next slide. Yeah. You are both muted and you are muted, Henry.

Oh, I know. I muted the other one now. Before we do that is going to be removed in the edit. I'm just saying. The we've had a question come in. It doesn't pop it up on the it looks like a really good app, but seriously, how does skin temperature come into the equation? And so just before we moved on slide, I just wanted to Yeah. So.

I can address that from the mathematical side and from the clinical side. From the clinical side, believe it or not, there was one clinical trial showing that an effective treatment would modify your skin temperature, but I'm not really using it on the clinical level. And and currently, we're again, we're using Samsung. We don't have skin temperature. The goal the I mean, the clinical goal is obviously to use markers which are relevant for a d h d two a ADHD ADHD treatment. Obviously, cardio, which is kind of treatment against stimulants. Obviously, movements, galvanic skin response, sweat is an important element in analyzing ADHD. But mathematically, as I mentioned, even even if there is no, like, direct.

Clinical usability to a specific sensor when skin temperature is a a a good example, we are still using this this data. For example, we're using the blood volume pulse, which, again, it's not there's no, like, real clinic clinical essence in looking at it initially. But if this marker is actually changing under the medication, statistically, it is changing. It's actually helping us in doing separations, we will use it. So, specifically, the skin temperature wasn't that crucial of a marker. Again, any marker can only assist in the separation. If if the marker will not assist in the separation, the machine learning will simply ignore it. But you are right. Meaning, skin temperature is not the initial marker that we look at standalone.

While while analyzing ADHD patients. The movement is a bit more relevant.

And I thought that what was interesting, I'm just gonna bring up your slide here, was that you used one of the things by using all these sensors, you can see a change with the medication coming off and coming off and on. And so and from that, you infer the the the during that period, that's when there's going to be the impact. And that's so that's a part of what you get is to know that medication is active at that time.

Precisely. And how how valid statistically the separation was. If if you would like, I could just rush through the again, I don't want to bother you. I know that we are almost out of time.

Boom. Amazing slide.

Yeah. It is. Again, it's it's not I mean, this slide is is not associated with the gold standard. At stage at that stage, like, intermediate stage, we simply wanted to see whether we can automatically separate the off meds and the on meds twenty minutes measurements before switching it to the gold standard. So as you can see, all in all, it's it's it was very clear. Even though there are, like, see there are, like, six cases in which the separations were kind of lame, I I guess there there might be a possibility that that these are no no responders because, again, it's off meds versus on meds. If you can move to the next slide. And maybe another the technical slide, if you can move to the next one.

Yeah. This is actually the the slide that sums up what we did in Israel because here you see on the x scale, the twenty minutes off meds and the twenty minutes on meds. And on the y scale, you see all of the kids, all of the patients, the 37 patients. And, technically, we try to use, again, the machine learning to predict for each and every thirty second time slot by training it to mimic what the physician said, by looking at the DSM five compliant data that we got from the physician, whether the machine sees patient off meds or on meds. And even though which was only the preliminary dataset and, again, it was, like, two years ago, I mean, back there, it was back then, it was very easy to see that even though it's it's a tough assignment.

Again, try to mimic the bleaching nature of ADHD, the sensors are exploding with a DSM five compliant data. I will state it as a decision support. Again, I I I think I mentioned that we don't aim to automate.

What what do you mean when you say DSM five compliant data?

So, basically, the quantification done by a clinician when the clinician is actually analyzing the patient, you know, one to nine scale, have a back above six six symptoms, meaning positive for ADHD or may maybe the kid is under medication, but the medication not effective or below six symptoms. So, technically, this is what we tried over. This is actually the data input that we used to train the algorithms on. But, again, on the binary level, it's not this is eight and this is five. And then bad versus good.

Okay. We had a question coming in, which was that in my observations, only on and off medication, the only difference has been increased heart rate and sweating. No change in no change in movement. And now you specifically have measured movement using the device. Right?

Yeah. I mean, we're using, again, an ensemble of sensors including movement. Actually and I think that was the question. Actually, it's even though the movement is an important tool or market or attributes, even while you are using the questionnaire, you know, there are some movement related questions. Looking at it, again, statistically and taking the movement as a stand alone marker, it it's it's it's not easy. It's pretty difficult. I mean, if you look at the movements from the watch, I mean, accelerometer as an average figure.

And I put your slide up with movement in it. Yeah.

So so you see, like, a slight decrease, but, again, it's not it's not valid. But, again, the machine learning is not looking only at a simple analysis whether the figure is higher or lower. It's a bit more complicated than that. I showed you, like, in the slide in the middle, whether the data is consistent or not, meaning whether at the end of the day, the movement is, like, unified or or streamlined or or broken or not. So, again, it's machine learning. So it's not as if we're looking at the heart rate and looking, okay. It went up. Medication is working. We're looking at the movement. It went down. The medication is not working. By taking machine learning, obviously, it's not four sensors or five. It's 500.

It's standard deviations, relations, correlations. So.

What about for people who are trying to understand? Say, look. You you've you've measured that their medication is having an impact on the body. You we talked about the the story, which I think you've like, I like, I've been coy on, but it was your was your son about where medication had impacted hyperactivity and behavior but not learning. How do you measure those aspects? So you've you've got a measurement saying, right, this is where medication is coming off. What what does that then let you do? Yeah. So.

At this level, at this stage, we're simplifying it. For example, let let me address your question from a different angle. For for example, the side effect. If the medication gives or if the medical treatment has side effect, for us, it's a constraint. It's not something that we actually incorporate into the models. Meaning, we don't calculate, we don't predict whether under at least at this stage, whether under the medication will will the the child will have sleeping problems even though we are measuring the duration. So I I don't think it's a good example, but, you know, stomach problems or some other issues for us.

We are doing something which is pretty straightforward. We are telling you whether we see a good separation or not, and this is the intraday pattern of the separation. If there is a constraint in which for ILUVIA, the medication is defective because we see a strong separation, but under this medication, the kid has stomachache, so it's a constraint. So it's basically some sort of an off balance sheet solution. So we are at this stage, we're not incorporating other elements such as side effects, such as behavioral issues even though, again, there is a strong correlation between effective stimulant treatment and and behavioral issues. We're trying to keep it simple.

Whether the impact of the medication is significant or not. And together with that, looking at the raw data, whether the impact was too strong for on the cardio, for example, or not. Yeah. And, actually, you jumped to the clinical trial that we did in Cincinnati. So as I mentioned, they told us at the get go. For us, the holy grail is not only knowing whether the medication is working or not. In some cases, they are, like, using dual dosage, like like, intraday medication. So it's a bit more complicated. And as I mentioned, they guided us through the goal of actually doing the analysis on an average day.

Again, by admin, it's not like an one day measurement because, again, we have to clean data. We have to rule out cases in which the patient was playing basketball at the playground. And so but but, again, we're using a lot of data to portray the average intraday pattern. Again, it's not as if we are measuring the drug level in your blood cycle, but we are measuring how the how much how statistically significant was the overall change in your physiology. And here, again, I mean, it wasn't an easy assignment, but but as you can see, the I mean, now it's working pretty smoothly. I mean, in some cases, it's not working, but maybe the medication is not effective. But, typically, we do see I mean, we always see patterns.

The main question is how what is the height of the pattern, meaning whether it's statistically significant or maybe the change is not significant enough, and then maybe the the physician should try and examine it and maybe titrate to calibrate.

Thank you. I wanted to show those graphs because I think they're fascinating in terms of the that intraday impact that you're you're able to measure that sort of detail. What's the impact? Like, look. So right now, we're in clinical trial, but we've got the watches on our wrist. Like, so what's what's oh, I don't. Although I'm seriously thinking about anyone. What is the for us right now, for adults, like adults and parents, should we be thinking about now?

So, again, obviously, the parents are a crucial gatekeeper a crucial stakeholder because, again, I think we are the gatekeepers, and, again, I established with my wife as as parents. But I think you should wait for the specifically for you for the NHS to adopt it because I don't see us going direct to consumer. We are examining the special education ecosystem as a nonmedical device. But.

As we see, the main growth That's that's interest just so people can know about that. So what what's happening like, is in special education, you're looking to help the children with ADHD manage have impactful medication throughout the school day. And schools wanting to impact as a part of their package in supporting children are seeing your tool and going, this could really help. We would know that this child is metabolizing this really, really quickly, and this child is not. And that will help us really help them. Yeah. Is that about right? Have I got that right? It's definitely right. Basically, the the the overall.

Mission is is kind of similar to the medical version, but the I think the main difference is the statement. Because in the clinical trials, we show that we can actually mimic the, as I call it, the DSM five compliant binary analysis or providing for the patterns. While working with, I would say, nonmedical version may be schools or employers or I think special education schools. I mean, by nature, we're giving something which is decision support. I mean, even if we are mistaken, we don't see an effective treatment. But with school, it will be a bit of a lower level in the sense that we don't see any physiological change.

Under the medication. Now it should go to a physician. The physician should decide what to do or not. And, again, it might not only be the case. I mentioned my, again, my young son case in which we actually avoided going to the physician because we saw still a strong separation. And before bothering the physician, we actually tried to look for an alternative solution, which was moving the child to another place in the class, which surprisingly worked amazing on the same day. So it's it's it's again, it's shedding light into the ecosystem. It's it's giving more objective data points. But, again, I should emphasize it. We are not automating anything. We are not replacing any analysis like this in fact compliant analysis processes.

We're just here to shed light and to provide objective data to this super subjective manual trial in our process.

Fantastic. We're out of time. It's yeah. That's been so interesting. And I I thank you for sharing all the data, sharing that information. It'll be for many, as it was for me, the first time they've seen that sort of thing and just absolutely fascinating. And we've answered some of the questions as they've come in, which is sort of what I've tried to do. We had an interesting thing happen over the weekend. We had the deputy chairperson of the conservative party talk about that there are parents with five children. They've all got a diagnosis of ADHD, and therefore and they're using it to defeat the benefit cap and that they're actually just bad parents. Now.

In your introduction, you mentioned you have four children, all of whom now have a diagnosis of ADHD. So what would your response be to How can I chairperson?

Know, ADHD, I mean, technically, it's a disorder. But, honestly and I'm not just saying that. I'm trying to be honest about saying that. I I do see it as an attribute. I mean, we know we all know. I mean, the the the physician that spoke before me, it's it's like dopamine. I mean, everyone knows. And it's very easy to blame the kids, the parents in bad parenting. The child is misbehaving. He's not polite. I I don't even know where to start by addressing that, but I will say if I will look at the bright side of the moon, we should I I think that our goal as parents is to see how we can actually harness this attribute and utilize this attribute. Because, again, looking at ADHD, at adults with ADHD, and I think we talked about it.

I think I got this analysis from UK, The UK, if I recall. There there's almost no middle. You will see many entrepreneurs with ADHD, many physicians with ADHD in the army. I mean, certainly, army, the army loves ADHD people. But together with that, you have the dark side of the moon. If the disorder is not properly treated, you know, the prevalence among prisoners, unnatural death rates, thirteen years less life expectancy and and and so on and so forth. So I think that our role, and if I may ignore this assumption that we're simply bad parents, I don't think it's the case. I think our role is is to navigate the child while having this attribute and to utilize it rather than suffering from it.

Thank you. You've been amazing. We have to say goodbye now and go to the next. I found it fascinating, genuinely. And I it's I think what you're doing is incredibly interesting, and I look forward to further updates. And as you said, you're doing part of a very big trial within the NHS. So this is something that's becoming towards people, hopefully. At which point, I have no doubt we will get a load of press talking about how ADHDers are just like getting free smartwatches, and it's just all some kind of gambit to acquire a watch. But wonderful. Thank you so much. Thank you for your time. I really, hugely appreciate it. Thank you, and have a great have a great rest of your day. Thank you so much. From now. It was a pleasure. Goodbye.