Brainnovation: Decoding the human brain (ep25)
Nemanja: This is Chair, a place where we discuss innovations.
The human brain is still a very big mystery to us even though it’s the 21st century. Today’s innovation or, should I say, brainnovation takes on the challenges of integrating technology into the human body. The brainiac behind one such idea is Ivan Gligorijević. Ivan is currently CEO and co-founder of mBrainTrain, a tech startup focused on high-quality variable EEG systems. After finishing his PhD studies in biomedical engineering, he embarked on the journey to decipher the human brain. So, Ivan, welcome to Chair, a pleasure to have you here today.
Ivan: It’s my pleasure to be here!
Nemanja: So, in the beginning, what was the motivation behind combining biomedicine and EEG systems?
Ivan: Well, there are various answers to that question but I will be very honest. The motivation was the opportunity because, as you said, biomed — the Group where I completed my PhD in Belgium was called biomed group, and it dealt with the physiological data and processing them, but the very name suggests that the context in which this was observed was medical somehow. And why was that? Because, to reliably observe physiological data, to record physiological data from the human body regardless of the type of data, you had to have some setups that were more medical than anything else, so bulky in isolated rooms…
N: With controlled conditions and everything.
I: Exactly. And the issue there was at least for several questions, there was no efficient way to address them. For instance, if we are talking about our brain, can you say that somebody was relaxed in natural condition if you instructed him or her to sit, you know, still look upfront, not move and be relaxed.
N: So that’s not natural at all.
I: That’s not natural at all. And it turned out that several questions could not be addressed at all. So, for instance, one major thing that could not be addressed is how we interact on social occasions. Like, now we’re talking, is my brain the same when I’m talking to you and we have this interaction and when I’m just doing something else or observing some content? So, there was a consensus that there were some questions that could not be answered, on one hand, and the other hand, there was technology. So, I say this may be often, I don’t know if I said it anywhere public, but, basically, when I came to Belgium I took my Nokia phone, you could now call it a “stupid” phone, but back then it was just a phone. And by the time I left, I was carrying my Samsung touchscreen phone in my pocket, and the revolution that happened there was that there was an immense amount of computational power in my pocket.
We all carry computers in our pockets. And together with the infrastructure that supports it, like, all kinds of first 1G, 2G, 3G, now 5G networks, this opened an opportunity to do all kinds of things. And, on the other hand, there was electronics an advancement in electronics, and we realized that a setup that used to take half of this room could be packed in a box like this and placed behind the hat. So, it still probably looks a bit weird to a lot of people having a sort of Water Polo cap on your head, but, let’s just call it, start. But this enabled all of these questions or at least some of them to begin to get their answers. So we saw that opportunity and we were fascinated by it, honestly. And, you know, the very thing that you can be the first in something and that you can be a leader in that, that was really fascinating. We didn’t think much honestly about all those business things that they advised, like, talk to the market, observe the need. We were just focused on, you know, let’s just do it.
N: And you haven’t told me that, who is behind mBrainTrain?
I: Well behind the mBrainTrain is professor Maarten De Vos from Belgium, professor Dejan Popovic from Serbia, Dr Bogdan Mijovic, also my colleague from biomed group, and myself. We are founders.
N: So, I want to get deeper into the innovation that you guys did, but since the principles of electroencephalogram… have I said it right? Okay! …are not so well known in the outside field of medicine. Can you explain to me and our viewers what exactly is EEG and how does it work?
I: Yes, sure! You know that our brain is composed of neurons and that somehow together and connected in some still very unknown way to us, functionally unknown, this creates who we are basically. And you know that there are layers of the brain, you must have heard about the reptilian brain, and so forth. But there is something very distinctive in our brain compared to our, let’s say, animal relatives, and what we know is that what the very distinctive thing is, let’s say, focused on the most revolutionary new part of our brain, is the neocortex, that’s like just on top. In this very thin layer of the brain, it contains most of who we are, and what we do, and how we reason, and our memories, and how we talk, and so on. It turns out that when these neurons work they, let’s say, fire, we call it fire, fire these spike-like events, and on a single neural level this is very hard to observe. But together, these neurons, create these spikes and that’s how they communicate to one another. And if you place, let’s say, some kind of electrode near this neural tissue, you’re able to observe the difference in voltage and currents, but let’s stick to voltage, electrical voltage. And it turns out that if you place some electrodes on top of your brain, you could actually record some of these changes. Now, to be able to observe anything at all, a lot of neurons, beneath that electrode, have to be working very synchronously to add these very small, small amounts of potential difference to some observable level that you can pick up at the surface of your head. So, this is, you could call it, one very simple model of what really happens, so we are dealing with one very simplistic view of our brain. Like, imagine that you are looking at the picture and, instead of a high-density picture, you see one just so pixelated that you need to focus to see who’s on that picture at all. That’s the analogy that I use with the EEG.
N: So with this homework done we can move on to your innovation. And I would like you to share how it was conceived and how this innovation transformed over the years?
I: Yes, well, it was conceived with the help of a lot of beers in mostly German breweries and a couple of friends that share this idea. So basically, at the start, there were three of us, all former colleagues in this biomedical group that I mentioned, at KU Leuven in Belgium. One of us, let’s put it like that, was already in his next position and as a postdoc at the University of Oldenburg Germany, and it is the focus of research of professor Maarten De Vos that got us interested in what could be done. And then we started brainstorming. There was some proof of concept already being done at this lab with some duct tape prototypes, you know, playing around with what was available on the market. But we realized that if we could really integrate it and, you know, bring it to the next level because what was, you know, that duct tape prototype was not comparable to, say, research-grade equipment. It was way more simple, more limited. And that’s the start of the idea, brainstorming how we could get it to work. We proposed that we do it here in Serbia largely because we had some kind of network of people that we assumed would come in handy as collaborators. We got a grant from The Innovation Fund of the Republic of Serbia in 2012 and that’s how we started. Of course, it took us some time to get going to realize that we are a company and not, you know…
N: And then the questions started to arise about the business and the market in it.
I: Yes, those boring questions that really, you know, defocused us from our work, that we found really a menace, but, of course, we were wrong but that’s the learning curve. In the end, as you asked, the innovation was the very small box with some electronics in it that you hold behind your head with the electrode cap that communicates with your mobile phone or computer. And that allows you to go into, let’s say, different conditions in even a home environment and do a lot of things… So, of course, we had different ideas on what this should start with. One of these was brain rehabilitation following a stroke. This was the first set of users, who used this for research in this domain. Like, can you help someone to recover or partially recover from a stroke in some friendly conditions like the home? But then we realized that this actually extends to a lot more questions than this and since, of course, we didn’t have any real marketing, or real market PR strategy, or call it as you will, we relied on word of mouth. And most of our first users just were with some friends of the other users who said: “Okay, this is really cool, I need this… I want to do this or that… I want to do sports studies. I’m really interested in how top performers get in the zone” or something like that. And then, you know, a whole thing followed with the auditory research, with visual research, with the social studies, and so on.
N: So where are you today, what are the challenges today?
I: Well, I’m sure you heard people say: “small kids — small problems, big kids — big problems” — that is also the case for us. So, as you grow, there are a lot of things that you want to do, that for some reason you still cannot do. And you work towards getting there. We enhanced our device and solved a lot of challenges that early users had. We just recently launched our smarting pro device, which really is, I believe, the best device by far in the market worldwide for mobile EEG research. That is, let’s say, one direction. So we established ourselves as a supplier of top-notch equipment for research. But even more than that, let’s say that there is a track on how to utilize this technology further and make it useful for everyday people, or in challenges that life and work bring. So we had many technological obstacles that in research are actually not such an obstacle, like gelling your electrodes. There is a conductive gel that you put in these electrodes to make the contact better. That’s unacceptable in other conditions. We want our electronics or whatever we use to be very friendly and a no-brainer if you want to call it that. So we are addressing a lot of these challenges and we believe that such devices have a bright future in work and life that could enhance who you are, how you perform, help you be better, and be more fun.
N: And with that said, how do you think that your product is going to impact the world of medicine?
I: Well, I think in many ways. I have to say that our product is, although it is used in medical research, most of the users come from actually cognitive psychology. So answering questions about the healthy human brain rather than ill. However, in the world of medicine, there are a lot of questions that actually scream for an answer, call it like that. So for instance, coming from the most obvious ones, like epilepsy. You have some people that need to be accurately diagnosed, which is a very hard task in, still bulky, medical setups. Because in medicine, the main things go slower than in the others, for objective reasons of course, but if you could make this technology available for more people so that they can do some kind of pre-screening at home, where epilepsy is suspected, then this could bring a lot of benefits. I think one or two per cent of the world population suffers from epilepsy. Then, in even more unknown domains, like Alzheimer’s disease or dementia. This is also very hard to diagnose. We don’t know what exactly led to that.
As one neurologist friend, I could call him like that, told me when I asked him about a healthy human brain, he said: “Well, I don’t know how a healthy human brain works”. I was like: “How don’t you know? Like you are the authority in this domain”. And he said: ”Well, to tell you the truth, not many healthy people come to visit me. People come to visit me when something is often very terribly wrong with them”.
N: So his point of reference is not the best…
I: Exactly! So we realized that we don’t know what led to that and this is what we call biomarkers. So imagine when, before, somebody visited the doctor for such a thing, there was some kind of continuous recording that helped us identify these biomarkers. So what led to that, we could have a more timely reaction and maybe even reverse some of the processes we don’t know about, but the problem is that healthy people do not go and get their EEG done regularly. And our idea is if that could be coupled with something that you immediately need, some of your daily need, that could come as a side effect. So let’s imagine that we all have some kind of invisible or seamless or concealed way to record our EEG that makes our lives better, and as a side effect we have the ability to see and get some kind of warning when something is changing. That could be the real benefit. That’s the medical sector we’re talking about, but the other point that I also touched on in our everyday life — benefits that could come from that. Imagine. We would call that a third hand. Something that enhances your performance and not for the sake of overperforming like a robot but for the sake of being better, more fulfilled. Now, in today’s world, we live in a world of abundance. We do not worry like our grandfathers did, about what we are going to eat, or whether we are going to survive. We worry about other things. We started wanting to be more, to be better, to learn more, to be better at our work, to advance. To do all kinds of things, to relax more, to meditate. All of these kinds of things.
N: I wanna ask you about data since I’m sure that you’re collecting a lot of it from users of your system. How are you leveraging this in terms of changing the product or some other area?
I: Oh boy, that’s a very difficult question, because there is obviously an ethical component to that. A privacy component, so to tell you the truth, from most of our users we do not get access to their data. And it has to be that way. You have to be very transparent about what you do, of course, with some others, where we are contracted to do something with that data. Then this is a different point. We do collect a lot of data on our own in very carefully designed settings. This information era that we’re living in, the era of advanced algorithms that we are seeing in the domain of AI machine learning, helps us to actually find better algorithms for some things. Traditionally, algorithms, I’m not sure how much people know and how algorithms came to be. But basically, our human mind is tuned toward some kind of insights so you see some things and you see, you hypothesize what you should do to get somewhere. And that’s how algorithms came to be. We take some signals, we process them in some way with certain assumptions, and say — okay, when we do this and that, in this number of steps, we are going to have a certain feature. And this feature is going to correlate with a certain thing that we can observe. A behavioural thing, for instance. Then we test this and if we have a high correlation on that we say — okay, this feature is a reliable predictor or a reliable sign that this and that happens. But are these algorithms optimal? They’re not, they’re not at all. There is no way to tune this in some way, you know, let’s say to finely tune this. And the algorithms that we are seeing now, that you know, most of you or us have at least some insight to. So we’ve heard about, like deep learning AI and so on. We know that Google works in some way that knows us, that can assist us in some way. But these algorithms are tuned for specific purposes that are not really good for the types of signals that we are seeing with the EEG. So we are working on algorithms that take advantage of the tools we have but are conditioned to the purposes that we want to address. The real issue with the brain is very often not having a gold standard, and that is the real challenge.
N: It’s tricky.
I: Because our brain is a tricky organ in itself. If I asked you what was your best holiday ever, you would probably tell me that this was a perfect holiday and you were with a bunch of friends and every moment was perfect. But basically, your brain is probably lying to you. Because, yes, indeed, that was the best holiday that you have. But in this memory recall, every time you change something in your memory, it now appears to be a lot better than it actually was. So there is a bias even if you ask someone about something. This recall bias. So the only way to really know about things is just some kind of objective analysis at the moment when things are happening. But even this is tricky. If you watch a movie, and you say: “I was scared”. Were you really scared? How scared are you? Like even if I ask you to scale something, then this scale also might be biased. So you need to really carefully design data sets and experiment settings based on which you’re going to extract features from this AI approach. And be confident that these features are really reliable, so that the next time you record these features, yes, that will mean that you know you were scared of Freddy Krueger or whatever.
N: Okay, so earlier you told me how your invention, how your innovation to be precise, is influencing medicine. But I want to ask you from a different perspective, the personal perspective because lately, we hear a lot about mental well-being and how this affects work. Probably because of this crisis and everything that we have changed in the workplace and so on. So how does stuff like mental fitness and or music therapy really help people?
I: You know when you wake up and you know you’re going to have a bad day, right? But you don’t quantify it. You just can maybe remember after a few days, or say — okay, I was really good that day or not. Measuring whatever that is that you’re measuring gives you a perspective and gives you some tools to affect things. I’ve already mentioned that bias doesn’t allow us to see things that way. But you can see that already now, time is not the currency of work as it used to be. If you were producing bricks for houses then the time would be the measure that correlates with the output of your work.
N: And the number of bricks.
I: Yeah, the number of bricks. But in today’s world you see all the creative activities that happen, what can you measure there, right? So what do you do? Do you, you know, make your designer just log in when he comes to work (or she) and log out? Does it tell you anything? It doesn’t. And we realized that, not only us of course, the entire world. So the currency is something between motivation or effective work and actually, motivation and the well-being that lead you to this effective work. For that to happen you have to be conditioned well. You have to sleep well, you have to eat well, you have to have productive behaviour. All those things that we are often not in control of and that we disregard because we cannot measure these things, let’s say, very precisely. We cannot say or, at least, nobody thinks about that, how a burger that I ate today and yesterday and the day before will affect who I am tomorrow. That brings us to the space of how to optimize ourselves. So one of these things as you mentioned is music. We all know that music affects us in some way. We all know that we can focus on some playlists. We all know that we can calm down with something. We all know that if there is some repetitive task that we want to do, we better put some appropriate music for that. But this is all very empirical. I believe that the technology that we are working on can make this thing less empirical and more knowledgeable and data-driven. These things are highly individual, and there is no one-fit-all solution. If you and I were listening to the same type of music, the output would not be the same. Maybe I would be distracted by this music, and you would be a great performer, and advancing and whatever. In principle, we talk about productivity, we talk about stress, and with all the tools that we have in today’s world, we haven’t addressed these questions. Precisely for that reason because there is no one-fit-all solution.
Otherwise, we would all just read Novak Djokovic’s book on how to focus and we would all be focused.
N: We are going to have a great backhand right?
I: Yeah, exactly! We will be focused and we’ll say: “Oh this person, you know, this one person figured it out so let’s just apply his formula”. It doesn’t work like that, our individuality doesn’t allow us to do that. So physiology on the other hand, if you ask me before most EEG can help address these things. Primarily by screening you, your habits, what works or doesn’t work, and make some kind of informed loops and solutions that would actually guide you in the direction of where you wanna be, right? If you’re a worker in a factory, maybe alerting you to upcoming danger or telling you that you’re not in, let’s say, good mental shape to conduct the work you have been doing. For instance, flight controllers, right? You cannot have a hungover flight controller or pilot unless you’re Denzel Washington and you can land the plane perfectly. This is one layer of these things. Another layer is getting you to relax efficiently and assisting you in that way, and making your life more fun. So I believe that we are going to start seeing these wearables. Now one thing that you didn’t ask me but I’ll still say, why I think this is the way to go in physiology in general, but EEG specifically. We are surrounded by all kinds of things that have some kind of measure of our behaviour starting with our phone. What these systems lack is how your brain responds to such things directly. We can map your behaviour but we cannot map your responses. So we cannot say what fits you, what you like, when you were distracted, how distracted you were, how to make you less distracted and so on. So for these things, you have to measure the brain directly and you have to do it at the moment when something happens, not after that because of the brain bias of the recall bias that I spoke about. So that’s what I think that we are going to start seeing, these variables come today or in maybe five years, maybe even less.
N: We are coming to my question about the future. This industry that we are talking about right now is in its conception. How do you see it in, let’s say, 15 years from now, will we be living in the matrix at the end of this century?
I: You know, I also don’t like people watching my screen when I’m doing something. You know, I feel under pressure. Even on my screen. People I’m surrounded with, you could say, that even if they saw what I’m doing, that I’m slacking, they wouldn’t tell me. But I don’t think that these technologies should be invasive in the sense of control. I think that there is an ethical component to that, not ethical in a traditional sense, but ethical in the sense that we should aim to assist people, not to lock them in some kind of jail. And if you, for instance, look at the industry and all this lean production, pioneered by Toyota, I think that this lean methodology focused on the production and not on the humans… So it optimized production but it didn’t optimize the mental states that we face during those things… And I think this is becoming a topic now. So I don’t think that physiology here will be used to extract, to squeeze, you know, people even more, but to see where and how people react and adapt the other part of the equation to them and not vice versa. Because we clearly have limits, you know, not extending the story too much, but can you calculate as fast as your calculator can? Of course not, so there is a limit. There are many more but so and we cannot change ourselves there is still an irreplaceable component of our soul. So we have to adapt our surroundings to match us and our abilities. I believe that this is one part of the equation. The other part of the equation is the technology that develops on the other end. We have an industry where the majority of errors are human-related and not technology-related, but we, as humans, have to keep liability in our hands because you cannot say that AI was at fault that somebody died or something crashed. It’s not an acceptable answer. It can happen once or twice while you’re developing technology but it’s not an acceptable answer. We have to find a layer in between to efficiently communicate with the advancing technology. And I think this layer in between human physiology relating mostly to our mental states is a crucial step.
N: Ivan, thank you so much for this conversation. I’ve enjoyed it a lot. And, for you out there, thank you for watching. And subscribe if you haven’t already. See you next Thursday for some innovation talks.
I: Thank you.
Keep up with us:
