Personalized Brain Mapping Technology with Stephane Doyen, Co-Founder, Director & Chief Data Scientist of Omniscient.

(Intro Narrator at 00:00:03) Hello, my friends. Today, Joel is talking to Stephane, founder and chief data scientist at Omniscient. And they discuss how advanced personalized brain mapping technology is changing the game in neuroscience, how Omniscient is leveraging AI to create these brain maps, and the impact detailed brain maps are having on patients, doctors, and researchers. All of this right here, right now, on the Modern CTO podcast.

(Stephane at 00:00:33) Here we go.

(Joel Beasley at 00:00:34) This is the Modern CTO podcast. So you guys are making brain maps. What are those?

(Stephane at 00:00:49) That's right. We're making brain maps. So there's a little bit of background about the brain, maybe, we can go over before jumping into what brain maps are. But your brain, if you look at it, looks very similar. Right? If you take a brain out of a skull and then you start cutting through it, what you're seeing from one part to another is awfully similar. So that's hard for a neurosurgeon to navigate. Obviously, there's landmarks that they know, but, you know, it's confusing. And so, essentially, when you consider the brain or your cortex and some of the deep structures, they're made of functional areas. So what are functional areas? They're small parts of your brain that do something specific, like tiny microprocessors, if you like. So some help you to see, some help you to hear, some help you to move your body, and so on and so forth. And they all organize into networks. You know, there's connections between different parts that allow you to speak, for instance, a language network. For speaking, you need to move your jaw, think about the words, listening to yourself talking for feedback, listening to whatever question you just got. So all of these things need to happen in your brain, and they're all organized in different parts of your brain. Now, like I said previously, when you look at the brain, you don't see that. Right? You don't see this part is for language, this part is, you know, moving your jaw, and so on and so forth. So you need a way to tell that, and this is what we're making. We take a brain scan, so it uses an MRI, pretty standard MRI, if you like. And then we pass it through an analysis pipeline, set of machine learning steps. And what comes out on the other hand is basically a map of your brain, which tells where those different things are located and the network, the white matter pathways that is connecting them. That's what we do.

(Joel Beasley at 00:02:50) How do I get one?

(Stephane at 00:02:52) Well, you can turn to your local MRI machine and get it done there. I mean, it's really working with most of the standard MRIs, and that's one of the important things, you know, when we built this into a software, we wanted to make the software as available as possible for doctors around the globe. We wanted this thing to run not only on the great data you're getting out of universities, but we wanted this thing to work on, you know, whatever local practice MRI you have. So if you go to see a doctor and you get an MRI prescribed, you can run our protocol that will give you the data and we can generate your brain map.

(Joel Beasley at 00:03:33) That's pretty cool. So you can do it with existing MRIs. You don't have to know this before you're getting the MRI?

(Stephane at 00:03:39) Correct. Correct. So MRIs can run different ways. Right? So it's a big magnet that's pulsing around the brain, and you can make that magnet pulse in different ways. So you just need to enable different ways this magnet is pulsing, but you don't need a hardware upgrade. It runs on current existing hardware.

(Joel Beasley at 00:03:59) And then how is the doctor using this map? Are they wearing goggles and it's three-D over their brain and they can see where they're going? Or is it like I give you a piece of paper and, like, MapQuest back in the nineties. Here's the directions. Like, how does it work for the doctor? How do they interact with it?

(Stephane at 00:04:15) Yeah. That's the future state. So currently, what neurosurgeons are using is, you know, what they call a guidance system. So you get into the operating theater. I mean, we're talking about neurosurgeons here. There's other doctors that can use our map, and we can certainly talk about the future uses of our maps. But for neurosurgeons, really, you know, they've got this guidance system, so the patient's head gets put on a frame, if you like. And that frame is connected to some kind of 3D reference system in the room. And then our map just sits in there. So they've got their guidance system. There's a screen on the one hand, the patient on the other. And then they can see our map overlaid on the patient scan. They can see on the screen now. Those maps are created in spaces, in volumetric 3D spaces. So if you have a system that can overlay it on patient's brain, then you could do that. We haven't seen any guidance system that does it just yet, but it's something that's possible. Yeah.

(Joel Beasley at 00:05:16) That is so cool. That's the future, right, when you can just have the brain open and you can see all the information projected directly onto the brain.

(Stephane at 00:05:23) Yeah. It's critical, you know, and I don't recommend you to get any brain surgery, but some people have to. Unfortunately, this is what happens. Right? And so, I mean, we're working with a lot of doctors, neurosurgeons. You know, my co-founder is a neurosurgeon himself. And, you know, this is the kind of thing that was missing in the field, where doctors were making certain cuts in the brain, and patients would wake up with terrible deficits. And then sometimes they would make another cut, and the patient would have little to no deficit. And they were really struggling to understand, you know, what's going on? Why is this patient doing okay while this one isn't? They were just missing the information. And that's what this map is bringing is it's telling surgeons, hey, look at this network. You might be cutting into those fibers. And if you do that, the patient will have troubles when the patient wakes up. So it's a very necessary tool for neurosurgery.

(Joel Beasley at 00:06:24) So it's approximate in a sense. So you put them under the MRI. You can scan them to show essentially what's, like, one layer below the surface of what the neurosurgeon can see. Like, you can see all the way through. Right? And then you can give them a map because everyone's brain's slightly different. Right? So everyone's pathways have grown slightly different, and there can be clusters in slightly different spots. But there is enough information to know that in this region, generally, this type of thing is processed.

(Stephane at 00:06:53) Yeah. You're hitting on the spot. That's it right there. It's an information problem. So these, you know, scans are billions of data points. They're all there. But you give that to a doctor, give them a lifetime to work through all those data points, they wouldn't be able to make the answer. Right? And that's kind of what the latest advances in machine learning allow is, you know, making sense of those data points. So it's finding patterns. And so the thing we do is, you know, you've got your brain, you've got the gray matter that's outside the cortex, and then in between, you have the white matter, which are, you know, pathways that are connecting things together. And so you can look at those pathways, you can see how things are connected. You can use that information with, you know, a bit of machine learning there to make sense of it and build those maps. And what you're doing with that is you're creating something that accounts for individual differences. And that's particularly critical in the field of neurosurgery because there are maps out there that exist so you can get it done in universities and so on and so forth. It's painstaking. It takes a long time. We've automated it. We've made it quicker. But also, we've made it in a way that if your brain is not normally shaped, so let's say you've had a resection or you have a tumor, it actually works. Whereas other maps you'd get aside from taking a long time, the system would just crash and you would never get your map. So that's another, you know, aspect aside from the individual differences is also, you know, those kind of major differences that you see in medicine.

(Joel Beasley at 00:08:32) Now my stepmom and my brother are both physicians. So I grew up around, you know, talking to doctors. And they always talk about malpractice insurance, right, because it's expensive. And so my thought when I hear about this is, obviously, you're taking this product to market. Right? Have you guys explored the idea of going to the malpractice insurance providers and having them make this as something that gives them a discount or is mandatory if they use this? Because to me, it sounds insane that there could be technology that could prevent you from cutting into someone and having them lose some major function. And if that exists in the market today, how is it not a requirement or at least a discount if you use it?

(Stephane at 00:09:12) Yeah. Well, you know, I'm more a technician here. I'm not the medical system legal expert.

(Joel Beasley at 00:09:20) But, you know everything, Stephane. Alright.

(Stephane at 00:09:23) Trying to be. No. But things take time. Right? I think the point here is, you know, we're bringing information that wasn't available before. That's kind of where we are. You know, the use of that information is for doctors. And now where this sits, I think it's something for the medical system to figure out. You know, every time you bring new information, you're disrupting the status quo. Right? There was nothing and suddenly you're saying something and that something is meaningful and useful to whatever field. So that's for the medical system to figure out. The best thing we do is we make sure that the information we bring is, you know, the highest quality information there can be.

(Joel Beasley at 00:10:07) So I actually got to talk to this interesting company, Synchron, and they make this sort of electronic stent that they put up to the jugular and through the blood vessels in the brain, and then they leave it in there, and they built their own operating system, and so the person can control this operating system with their thoughts. But yours specifically works with an MRI. Have you collaborated with any other companies on trying to make your thing real time by any means? Is there any way to make what you do real time?

(Stephane at 00:10:38) Yeah. So currently, you're requiring MRI. The reason for this is that we're working on functions. Right? So we're working on how the brain works rather than what is where. I think that's the important part is the key to better recovering post surgery is to watch for the function, not the anatomy. So you still require an MRI nowadays. We're exploring other ways where, you know, you could probably get an EEG done or you could probably get those other technologies like, you know, near infrared spectroscopy, things like that you could use that would make it more online real time, but so far it's MRI only. Now the thing we further do with the maps, right, is we can look at the brain in action. And I think that's the next venture for us when you're thinking about activity. You know, once you have a map, you can describe things consistently. And if you get consistent descriptions, then you can start quantifying them. So let's look at your language network. Well, is it functioning well? Is it functioning too much? Is it not functioning enough? Let's look at the network that underlies your mood. Does that explain why you're feeling up or down? And does that actually trigger some potential helpful way forward? These kind of things. So while we don't do things online, you look at the brain in action, and that's an important property of the, you know, the future product line that we're looking at.

(Joel Beasley at 00:12:00) So when I was reading, I was really excited about the brain maps. You know, they sound really cool, but then they can be used for finding solutions to brain related disorders like depression, chronic pain, cancer, bipolar disorder, Alzheimer's, dementia, PTSD. Can you see if you're monitoring it in real time, can you see someone being depressed? Can you see that?

(Stephane at 00:12:22) Yeah. So the monitoring in real time part is the thing we don't do, and it's a sensor challenge. Right? It's probably something we can do, but it's gonna take a little bit of time to get to good sensors. MRI is what is giving you, you know, the best compromise between spatial resolution and time resolution. For these kinds of things, you don't need to be hypergranular in time. You need to be pretty granular in space. And so MRI is pretty good for that. And so we do some of that indeed. That's kind of the next thing we're looking at. So there's an interesting thing going on in your brain when it starts working is that a functional area will start to ask for more blood as it starts to work. And that request for more blood essentially gives you more blood in that area. And in blood, there's a little bit of iron, so you can pick that up with a magnet. And you get what's called a BOLD, which is blood oxygen level dependent response. And so you can see which part of the brain turns on and which part of the brain turns off as you do a scan. So we don't do that online per se, but you can get a recording of your brain. So you get into an MRI, you sit in there for eight to ten minutes, and then your brain in resting state, so when you idle, it never stops working. Right? There's always a background noise, and all the parts of the brain are kind of talking to one another, and networks are kind of working together. And so this creates this signal. And inside of this signal, you know, there's a lot of information. Right? So we haven't found anything outside of anyone's brain that will render thoughts, fears, memories, you know, whether you're hungry, whether you're awake or anything else. It's all happening in your brain. So if it's all happening in your brain and you can get a recording given that your recording is of a reasonable quality, you can start picking up these things. And that's a data problem, and that's what we're doing here at Omniscient.

(Joel Beasley at 00:14:29) That is so cool. Has it gotten to the point where you can put people in an MRI and just ask some questions for the pure sake of exploratory figuring out? Like, are you hungry? Yes or no? Or things like that. Are they exploring like that, or is it only when people have a serious medical condition?

(Stephane at 00:14:44) So the way we work currently is we're doing resting state. So it's really about how your brain functions in idle mode. You can do what's called task based. So you do ask questions or you say, can you tap your fingers? Or the thing with that is you're limited to what you can explore. So for instance, if you wanna test somebody's mood, right, can you on command be sad or happy, right? That doesn't quite work out. So you're gonna need to find a system that does work for that, and resting state is good for this. We don't ask questions, but we do measure, we can record, if you like, people's brain activity, and then can tell whether some parts of their network are not firing correctly. And so these networks are organized in functional systems, right? So you can say, okay, one thing we've done for instance is that we've explored patients with schizophrenia. And then we looked at, for instance, where in the brain was located some of their symptoms. So that's another, you know, important thing of these conditions is that, you know, mental illness is one gigantic, if you like, category. But if you consider it further, you can break it down into, you know, smaller categories. Right? So I mentioned schizophrenia, but look at depression, for instance, for a moment. There are depressions where, you know, people wake up too early, or people have troubles falling asleep, or people wake up in the middle of the night. There's depression where people, you know, lose weight or gain weight or, you know, there's all sorts of different permutations of symptoms. And so all of these symptoms have a different expression in the brain, and that's what we can pick up. So to my schizophrenia example, we looked at patients hearing voices, for instance.

(Stefan at 00:16:25) And so we use the same technique with the BOLD resting state, and we looked at where in the brain was located that specific symptom. And what we found is actually coming from part of the auditory area, which is not something we pre-wired. It's purely machine learning driven. So we basically had the scans, patients reporting symptoms, and then we looked at basically what functional areas were overworking in some spaces, and that's what came out of it. So we can look into these things.

(Joel Beasley at 00:16:57) That is so cool. Dude, this is the future. This is so amazing what you guys are building. When you see things like Neuralink, what Elon Musk is doing, are you all just considered in the same industry? Are there collaborations happening? How does the industry see its peers that are adjacent but not necessarily direct competitors?

(Stefan at 00:17:17) Yeah. There's a whole chain of things you can do on the brain. So one way to structure it very, very simply, thinking about the industries, you've got the read and the write. And so Neuralink is more on the write side. Right? What can you reprogram in the brain? Where Omniscient sits is in the middle. So we get data from sensors, MRI currently, looking at other potential sensors. And then we make sense of that information so that people that are on the write side—and so brain implants is one thing, there's other ways of stimulating your brain. Right? There's a thing called transcranial magnetic stimulation, which we have a lot of collaboration on. You can start reprogramming parts of your brain. But given that your brain is organizing functional areas and it's a complex system that's dynamic, how do you know which part to touch? Right? And this is what we're offering. These maps and these functional maps tell you, well, if you want to put an implant there or most so if you want to start, you know, looking into reducing those type of symptoms, maybe you should consider that area. There's some data in the scans showing that this might be the responsible of certain symptoms. And if you act upon it, you may change the course of action.

(Joel Beasley at 00:18:31) As the whole industry has been making these new technologies and maps and all of that over the past, you know, five, ten years, it's really advanced. Have we learned anything new about consciousness?

(Stefan at 00:18:42) That's a very interesting question. So I actually did my PhD in a lab called Cognition, Computation, and Consciousness. So it's a topic I did a lot of work on in my previous life. The thing at this stage, I should say, so there's some great advances on consciousness overall and trying to map it. Now you can break down the concept into, you know, many different bits. Right? So for instance, there is being conscious about your senses or your memory. And there's feeling conscious by hearing and seeing things, and that's called phenomenological consciousness. So there's many ways of looking at consciousness. Imagery is helping there. I should say for technology, this is a start. Right? So we de-skilled and we, you know, made access to that information so easy that researchers now can start, you know, answering those questions they have with data. Whereas before, it would take them months and months of preprocessing and probably wouldn't land on the same type of mapping and the same type of answers. So what I'm expecting is from systems like ours, we should see, you know, great advances coming out in the next couple of months and year, enabled by brain mapping and standardized quantification of the brain.

(Joel Beasley at 00:20:07) If you have to guess for fun, right, just for fun as a human, out having beers with your friends, what we end up finding out about—fast forward a hundred years and we very well understand consciousness. What do you think the explanation would be?

(Stefan at 00:20:23) Right. That's, we're getting deep here. That's where the beer factor comes in. You know, I think one frame of the debate could be, you know, does my conscious guide my action, my consciousness? Does my consciousness guide my actions? Or, you know, is conscious awareness something that comes at the end of the line when you've already done anything? And you find a lot of interesting evidence in the literature if you look at consciousness, and there seems to be a lot of it pointing out towards the latter rather than the former, that consciousness is a strange element of evolution that has arrived at some point with systems like ourselves. It's probably an emergent property of the complexity of our brain. And with that, we can find that action can be, you know, you could have your entire life led without consciousness. And so we're finding in somehow that it's more of an output reading rather than input generation. Now, why is this so? It's still unknown. Right? And of course, the question is still not as clear as I'm making it here, but ultimately, we'll find out.

(Joel Beasley at 00:21:42) That is so cool. We're pretty confident it's not like one specific tiny little section in the brain. We're pretty confident that it's more like a field or it's more ambiguous than just one location. Here's the point of consciousness.

(Stefan at 00:21:56) Yeah. So that's where you break down consciousness in various bits. Right? So if you think about the functional areas, so back to our maps. So different part of your brain does different things. And so part of your conscious experience is recruiting those areas or is actually working with those areas. Right now, you're sitting in front of a camera, there's a screen, you're hearing my voice. You feel like you're kind of watching a movie, which is the movie of your life. That recruits large areas of your brain, you know, and there's some theories, you know, you can look into the global workspace theory, for instance, which says consciousness emerges when all these things kind of work together. Now, if you look at, you know, function of the brain, we also found that when you're showing something and then you can measure the activity of the brain, you see larger activities that are more diffuse. The conscious experience is higher than if you've got only, you know, small parts of your brain that are just functioning when they're seeing things. Now that being said, there's also studies showing, especially in primates, that certain specific area in your brain when stimulated can either completely disrupt consciousness or bringing primates, I think the study was made on, back from anesthetics. So they're completely asleep, you stimulate that area, they wake up immediately, if you like. So there's a bit of both in there. Right? There's the conscious experience, which is, well, how everything is working together and how you get, you know, the film of your life. And then there is the conscious awareness, which might be relying upon specific areas there. And so we do a lot of collaborations. Right? So that's one of the questions we kind of explore with different labs around the world. Like I said, you know, the information we're bringing is completely novel. Researchers never had those tools before, and so we're helping them solving their own problems. So we don't solve them for them, but we do bring that information to them to the best of our skills.

(Joel Beasley at 00:24:01) Have you seen this Google sentient article that's been out for about two days where the Google engineers claim that the AI has become sentient? Have you heard about this yet?

(Stefan at 00:24:14) I have. I have. It's a pretty interesting one. I mean, it's one of, you know, there's a lot of discussions about this in the community. A few things can be said here. I think, you know, you certainly know the Turing test. Right? We're talking to an agent. You don't know whether that agent is a human or a computer. And if the computer is able to fool you, then the Turing test is considered to be passed. I think it's a pretty interesting argument for an upgrade of the Turing test, you know, because obviously it looked like there was some, at least that piece of AI did pass the Turing test in this specific context. Now, I mean, it's a bigger topic. Right? So you're looking at AI. I think there's different things you can see here. So I think overall, where a lot of people are heading towards is, you know, general AI. Right? Once an AI has actually passed a Turing test, great. What's next? Right? Is this AI going to start having volition and things they want to change in the world and is ultimately going to, you know, impact us? I think we're not too far, but still quite far before anything like this happened. But more so, I think it's quite unlikely that humans and AI, if you arrive to that distance future, will compete for the same resources. You know, I think we might find ourselves wanting and looking at different things, and in the end, I don't see a Terminator or a Matrix future. I think where AI is currently more challenging today is in spaces where, you know, it's actually impacting you on a daily basis and you actually don't know about or you don't quite feel about. You know, for instance, your news feed on Facebook, if you're connected there, you scroll through your feed every morning, and it's showing you things that are going to prime you for, you know, next couple of hours or your day, and that is actually impacting your behavior. And that's the space of AI, I think, is far more, you know, complex to navigate and is immediately challenging. I think that Google AI sentient is a really nice debate, but not something to be, you know, on the top of immediately, at least.

(Joel Beasley at 00:26:29) Did you read the transcript?

(Stefan at 00:26:43) No. I did get some bits and pieces of the transcript, though, not the full transcript, but I did see that there were a couple of interesting bits and pieces in there.

(Joel Beasley at 00:26:43) I went and he had a Medium article, and I was planning on spending like two minutes just skimming through. I ended up reading the whole thing. It was like a twenty, thirty minute read. It blew my mind. Like, it was really kind of strange. And one of the things that stood out to me was not only did it act like it had a form of sentience. Right? It wasn't like an—I have small kids. I've got a four year old and a three year old and one on the way. Right? You know, they talk differently at different stages, and then they have different understanding of concepts. But this thing seemed philosophical in a sense. Like, it was having deep conversations and making creative stories. And for me, I don't know. I was fascinated by it because I have more questions. Well, I'm going to try to get the guy on the show to talk to him about it, because I'm curious when he was interacting with this AI, he was asking it questions, he was giving it answers. But I'm curious, like, from an engineering standpoint, how do you initiate that? Can you monitor it when it's not engaging with you, and can you actually see it thinking or processing data?

(Stefan at 00:27:43) Absolutely. You know, I'm not an expert in language networks in a sense or, you know, this kind of latest advance in building those massive networks. I think there's two things there, perhaps. One thing is emergence, and that's something that's quite interesting, is that you can get some new property with very small basic principles. That's what you find in deep learning. And that's what gets researchers surprised. Right? They're creating a system using some basic principle, and then that system starts to do things that wasn't wired or coded before. The second thing is, you know, this, and that leads to that second thing, is as humans, we tend to make assumptions and hypotheses. Right? So when we're hearing language network having a discourse about, you know, physics or, you know, something they want to do and express wills and wishes, we tend to believe and judge that at the back of that, there's an agent that is actually making those wills and wishes. But there's a version of that where you can actually make those wishes without being an agent at the back of it. You know? It's purely, you know, statistics. It's connecting different things that you've seen before in a way that you've learned that would be compelling to humans. And that might be two kind of interesting data points to consider in the current discussion, but I think, you know, the debate goes far beyond that. And I'm very excited to see what comes out of that because I believe general AI is going to be less problematic than some people let us to believe.

(Joel Beasley at 00:29:21) I was blown away when it was saying, hey, tell me a story about the way you feel or whatever. It was asking these really odd specific things, and it was giving these unbelievable stories that showed a real deep understanding of certain concepts. And if they hit that on a statistical thing, then I want it to give me lottery numbers.

(Stefan at 00:29:44) You know? But it speaks also to how your own brain works. Right? So we're wired to pick on certain things. There's consistency in your brain, and so we're wired to when you're seeing or hearing something, it creates an experience inside of you that triggers something further thinking. Again, you're talking to an agent. That shows that our brain is built in a way that's quite regular, and you can actually play with that and you can get it to work.

(Joel Beasley at 00:30:14) When we're talking about this concept of an agent, I have a general understanding of what you mean by being human. Right? Because I can talk to certain people, like, if you go talk to some people that have mental disorders or even myself when I'm focused or doing something and my wife's trying to talk to me and I'm kind of not being responsive or I'm kind of giving her half answers and half involved, she's like, hey, you're not even there. You're, like, zoned out or whatever. Can you dive deeper into what this, like, the difference between an agent and a non-agent?

(Stefan at 00:30:44) Right. So there's an important thing you mentioned here, which is, you know, the concept of mental illness. Right? And so, again, the brain is a biological organ. It is a system that was crafted by evolution to process information. And the way it's optimized itself, if you like, is to process information in a reasonably consistent way. So we're, and it's generally driven by constraints of your body. Right? You have eyes and ears and mouth. And so these things generate information, that information gets processed, and then you've got an environment you're interacting with. And so that kind of shapes up, if you like, your neural substrate to crystallize or to form connections that are effective for dealing with that environment. And so I think the agent component is a complex one to grasp. Right? So part of the agent component undertones that there is something deeper in a sense that there is within that system, an entity that is following its own purposes to interact with an environment and its own design. That's hard to prove, of course, and there's a lot of research going on there. But I think the more interesting side of that story is mental illness can appear and there is no wish to be depressed. Right? So if somebody is having those thoughts, it's not that, you know, they decided one day, hey, I want to be dark and, you know, I want to feel bad and have troubles in my life. It's sometimes, you know, what's happening there is the system is kind of, you know, some of the optimization has gone out of whack, if you like. We've got a system that has crystallized itself in a different way. And so you take away the agent of the equation and you look at the biology, which I think, you know, if you want to create concrete outcomes, treatments, and recovery, that's the more interesting part. We can obviously debate, you know, whether that entity exists or not. But the fact of the matter is that if you want to be able to make a difference in patients' health and mental health, you have to look at the biology. You have to look at the brain and the connection. And so that's kind of what we were doing. You know, I mentioned the BOLD signal and the map. Right? And the two work together. You can first, you map that system, and then you quantify it in a way that actually speaks to whatever symptom you have. And that's how you can make a difference.

(Stephane at 00:33:18) And, you know, interestingly, ultimately, you might say once the symptom clears, then you might say the personality has changed. Therefore, the agent is impacted. I mean, there's a bit of an interesting catch-22, but I leave that to philosophers.

(Joel Beasley at 00:33:33) No, for sure. So I'll share a personal story for me. So I'm 34, and I noticed after, like, 18 or whatever, I started to get depressed, right?

(Joel Beasley at 00:33:44) You know, it was better and worse, and I attributed it to being an entrepreneur. You know, just fight through it, do your best you can possibly do, and so on and so forth. But then as you start to get a family and more responsibilities and you're not just a single person anymore, it becomes more important for you to figure things out. And so I started tracking, writing down, because I was trying to figure out, is it my diet? Because I'm good most of the time, and then out of nowhere, it'll just be like three to six weeks of just depression, and then I'm right back at it. I'm completely fine.

(Stephane at 00:34:11) Yeah.

(Joel Beasley at 00:34:11) And so this was happening between two and three times a year, and it sucked because I didn't want to talk about it, you know? And because it's so looked down upon, it's like, I don't want to be a depressed guy. Like, 99% of my life, I'm not depressed. I just have these weird couple-week areas where it happens. And so I finally got frustrated enough to where I visited a couple doctors and paid them to run tests and do all this stuff.

(Joel Beasley at 00:34:35) And it came out to like, there's something that fires off in my brain, and they gave me a prescription, and it's not like a stimulating prescription. I don't know what it is. I would tell you if I did, but I can't remember off the top of my head. But it basically, since I started the prescription, it just doesn't happen. It's just like, you know, I get nervous. Like, when three months comes up, I'm like, oh, it's going to happen. It just doesn't happen. It just keeps going. So it's been that way for several years now. And, you know, it's hard because you don't want to talk about it. It does have a public stigma associated with it. But it's not like I was being depressed. It's not like I was making bad decisions or I was abusing something. It was just something that was a process that was reoccurring that a medication stopped.

(Joel Beasley at 00:35:15) So I was so grateful. There's people like you researching this stuff and figuring it out. Because without the technology, I would, I guess, just live with it the rest of my life, you know?

(Stephane at 00:35:24) Yeah. You know, it's an important observation you're making there, right? Is that more can be done about mental health, and the stigma of it is tremendous, right? So if you sprain your ankle and you start limping, you're not going to feel guilty about it. And people are going to see it, and they're just going to offer you a chair or, you know, they'll consider it. But if you, one morning, wake up, or over time, you start developing these random dark thoughts, it's a completely different game, right? You're going to start feeling guilty about it, and you're going to struggle being able to exchange that with others because kind of, you know, deep down there's a belief that you wanted that or it touches your personality.

(Stephane at 00:36:11) But, again, I think it's really, really important to consider the biology here. You know, there's very little difference as far as we know between you spraining an ankle and yourself having dark thoughts. It's just a failure of a biological system that we're equipped with. And obviously, the failure is more complex in the case of the brain. It's just hard to highlight.

(Stephane at 00:36:35) You know, again, we're not doctors and we're not providing treatments, but we do help doctors with understanding better the patients. And so often what we do when we've got those brain maps, you know, showing how brain networks are functioning, and we're able to show to some patients, well, yeah, no joke, you're feeling depressed. You know, your limbic system, which regulates some of your emotion, is completely deregulated. And that creates, you know, a major difference for some of the patients is that they understand that it's not something that they wanted or it's not something that they, you know, suddenly got bad or had gotten there because of their action. It's just that the system deregulates, and it can deregulate for so many reasons that aren't what you wanted it for.

(Stephane at 00:37:23) And I think that's an important role we have to play here as a company, as Omniscient, is essentially provide the best possible information to doctors so they can start, you know, looking into these things from the angle where it should be looked at. That is the biological system interacting with the rest of the world.

(Joel Beasley at 00:37:44) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email, [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.