The Open Loop with Chay Carter

BrainScope's Matt Adams discusses EEG-based AI platform for concussions, dementia, and stroke mimic differentiation

39 min Listen to the original
MedTechAIEEGConcussionDementiaStrokeSports MedicineDigital Health

Summary

Matt Adams outlines BrainScope's evolution from a single FDA-cleared algorithm for detecting brain bleeds into a multimodal AI platform integrating EEG, blood work, speech, imaging, and genetics. He highlights the platform's clinical utility in rural medicine and sports medicine, citing a 99% sensitivity rate for detecting brain bleeds down to 1cc of blood. Adams also discusses the company's pipeline for dementia biomarkers, which he claims can predict onset up to seven years earlier than standard care, and its ability to differentiate stroke mimics from actual strokes. The platform is positioned as an operating system for third-party developers, leveraging a deep learning model trained on 170,000 patients.

Episode sections

  1. 0:49 – 2:08

    Company History and Background

    Matt Adams outlines his 25-year medtech career at Medtronic and other companies before joining BrainScope. He notes BrainScope is a 20-year-old startup, not a new lab spinout, and highlights his arrival about a year ago.

  2. 2:08 – 3:17

    Platform Evolution and Core Technology

    Adams describes BrainScope as a multimodal AI software platform using EEG as the core measure, capable of ingesting blood work, speech, imaging, and genetics. He explains the transformation from a single FDA-cleared algorithm for detecting brain bleeds to a broader platform.

  3. 3:17 – 4:37

    FDA-Cleared Products and Clinical Utility

    The original algorithm is 99% sensitive for detecting brain bleeds down to 1cc of blood. The process takes 10 minutes total, requiring only 2 minutes of clean signal. It is designed for non-specialists and is used in rural medicine and field applications.

  4. 4:37 – 6:53

    Concussion Assessment and Return-to-Play

    BrainScope's definition of TBI now includes concussion detection, severity assessment, and recovery tracking. Unlike subjective tools, BrainScope provides a functional measure of brain activity to aid decisions on returning athletes or service members to duty.

  5. 6:53 – 8:46

    AI Development and Data Quality Challenges

    Adams states early development took 20 years because point-of-care EEG for generalists did not exist and required building data quality from scratch. He emphasizes that clinical research cycles to build high-quality algorithms for clinical confidence were lengthy.

  6. 8:46 – 13:16

    International Expansion and Sports Medicine

    BrainScope engaged with the Sports Doctors Network (founded by Real Madrid's CMO) and Vortex Performance Centers for MMA fighters. The goal is to improve athlete safety, accelerate return-to-play, and extend careers through objective assessment.

  7. 13:16 – 16:31

    Youth Sports Initiative

    Every Brain in the Game: Adams discusses a program targeting 12 million US youth athletes, focusing on underserved communities. The initiative aims to provide education and objective assessment tools, connecting identified injuries to sports medicine clinics for faster care.

  8. 16:31 – 31:12

    Dementia Biomarkers and Regulatory Pathway

    BrainScope has developed a dementia prediction biomarker published in Nature Scientific Reports, claiming to predict onset up to seven years earlier than standard care. The algorithm is validated against two external sources and is ~15 months from regulatory clearance.

  9. 31:12 – 32:40

    Stroke Mimic Differentiation and Disease Pipeline

    Matt Adams explains that BrainScope's published research demonstrates the ability to distinguish stroke mimics (such as severe migraines or syncope) from actual strokes in emergency departments. He notes that the platform is expanding its pipeline to include data for major depressive disorder, generalized anxiety, ADHD, and multiple sclerosis.

  10. 32:40 – 34:31

    Platform as an Operating System for Third-Party Developers

    Adams describes BrainScope's infrastructure as an 'operating system' allowing other companies to build their own biomarkers using BrainScope's proprietary data and sensors. He highlights a deep learning model trained on 170,000 patients across multiple diseases, which is currently in demo mode and attached to cloud-based AI like Claude.

  11. 34:31 – 39:19

    Multimodal Data Integration and Future Outlook

    The discussion covers the integration of EEG with blood biomarkers for dementia assessment, noting that while blood tests are improving, functional assessments via EEG remain distinct. Adams compares the future of neurotech to the cardiac world, where clinical, functional (EKG), and molecular assessments are combined to form a complete patient picture.

Claims worth verifying

Matt Adams has over 25 years of experience in medical technology, including leadership roles at Medtronic.

Unidentified speaker (UNKNOWN) ·1:12

How to verify: Matt Adams' LinkedIn profile and Medtronic press releases or employee records. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

BrainScope is a 20-year-old startup.

Unidentified speaker (UNKNOWN) ·2:39

How to verify: BrainScope corporate history page or SEC filings if public, or company press releases. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

The FDA-cleared algorithm for detecting brain bleeds is 99% sensitive down to a single CC of blood.

Unidentified speaker (UNKNOWN) ·3:17

How to verify: FDA 510(k) or De Novo summary documents for the specific BrainScope algorithm. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

The process from opening the package to collecting the signal to throwing everything away takes about 10 minutes.

Matt Adams ·3:37

How to verify: BrainScope product documentation or user manual. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

The algorithm requires only two minutes of actual clean signal from the brain to run.

Matt Adams ·3:57

How to verify: BrainScope product specifications or clinical validation study methodology. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

The original FDA-cleared AI-enabled product in the field of neurology was from BrainScope.

Matt Adams ·3:57

How to verify: FDA database search for 'first FDA-cleared AI-enabled product in neurology' or BrainScope press releases. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

There are 12 million youth athletes in the United States in concussion-oriented sports.

Matt Adams ·13:16

How to verify: CDC or National Federation of State High School Associations (NFHS) statistics on youth sports participation. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Early identification and quick care can improve concussion recovery times by 50 to 73%.

Unidentified speaker (UNKNOWN) ·14:33

How to verify: Peer-reviewed medical literature on concussion management and recovery timelines. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

In the US, 7.2 million people are diagnosed with dementia.

Unidentified speaker (UNKNOWN) ·16:48

How to verify: Alzheimer's Association 'Alzheimer's Disease Facts and Figures' report. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

The number of people at risk for dementia in the US is about 22 million.

Unidentified speaker (UNKNOWN) ·16:50

How to verify: Alzheimer's Association 'Alzheimer's Disease Facts and Figures' report. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

BrainScope has published research demonstrating the ability to separate stroke mimic from stroke patients.

Unidentified speaker (UNKNOWN) ·31:17

How to verify: Search academic journals for BrainScope publications on stroke mimic differentiation; check SEC filings if publicly traded or press releases for publication announcements. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Approximately 40% of patients in emergency departments with stroke-type symptoms are having a stroke mimic (e.g., severe migraine or syncope).

Matt Adams ·31:37

How to verify: Verify against medical literature or emergency department statistics on stroke mimics; compare with published BrainScope data or independent clinical studies. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

BrainScope has accumulated data for major depressive disorder, generalized anxiety, ADHD, and multiple sclerosis.

Matt Adams ·32:39

How to verify: Review BrainScope's website, white papers, or clinical trial registries for data collection efforts in these specific disease areas. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

The company has a deep learning model trained on 170,000 patients across multiple diseases.

Matt Adams ·33:48

How to verify: Check BrainScope's technical documentation, white papers, or investor presentations for dataset size and composition details. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

The deep learning model is currently attached to cloud-based AI like Claude.

Matt Adams ·33:48

How to verify: Verify via BrainScope's technical demos, API documentation, or press releases detailing partnerships with AI providers like Anthropic (Claude). For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

BrainScope is building a platform that acts as an 'operating system' for other companies to develop their own biomarkers.

Unidentified speaker (UNKNOWN) ·34:31

How to verify: Review BrainScope's business model disclosures, partnership announcements, or platform documentation for third-party developer access. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Blood biomarkers for dementia are making progress in identifying molecular pathologies indicating higher risk.

Unidentified speaker (UNKNOWN) ·34:39

How to verify: Search recent medical literature and clinical trials for blood-based biomarkers for dementia (e.g., p-tau, GFAP) and their diagnostic accuracy. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Ken Nelson has a strong background in the cardiac space and is working at Epitel.

Chay Carter ·38:24

How to verify: Verify via Ken Nelson's LinkedIn profile, Epitel's company page, or public announcements regarding his role. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify the FDA clearance status and specific indications for BrainScope's algorithms via the FDA database.
  • Check the publication of the dementia biomarker in Nature Scientific Reports and review the methodology for the 7-year prediction claim.
  • Review the clinical data supporting the 99% sensitivity rate for brain bleeds down to 1cc of blood.
  • Investigate the '40% stroke mimic' statistic cited by Adams and compare it with independent emergency department studies.
  • Assess the regulatory pathway and timeline for the dementia biomarker's clearance, noting the 15-month estimate.
  • Examine the data privacy and security protocols for the multimodal platform ingesting sensitive health data.
  • Look for conflicts of interest or funding sources for the Sports Doctors Network and Vortex Performance Centers engagements.
  • Review the technical specifications and limitations of the deep learning model trained on 170,000 patients.
Full transcriptexpand

0:01Chay Carter: Welcome to the Open Loop Podcast. It's a podcast exploring the people, companies, and ideas shaping the future of neurotechnology. I'm Che Carter, the founder of Carter Sciences, Recy Neuro, and the Neurotech newsletter. In each episode, I sit down with founders, operators, researchers, investors, and leaders building the edge of neuroscience, Science, neurotechnology, and healthcare. We'll have honest conversations about people's journeys into the neurotechnology. neurotech industry, what they're currently building, how it's going, the realities, the good, the bad, and the ugly, and what the future holds for the neurotech industry in general. Welcome to the Open Loop Podcast. I'm joined today by Matt Adams of Brainscope. Thank you so much for For joining me, Matt.

0:46Matt Adams: Jay, it's great to be here. Thank you for the opportunity.

0:49Chay Carter: Fantastic. Yeah, my pleasure. Um yeah, we've been trying to get this going for a little while now. We both had some surprise travel pop up in between, but really good to finally sit down. I think I I always like to start these conversations before we jump into what you're working on right now with a bit of bit of your history, because you're a 25-year medtech exec. Tell us about your background. How did you get into uh medical devices, let alone neuro?

1:12Unidentified speaker (UNKNOWN): Yeah,

1:12Matt Adams: Jay, as you mentioned, it's been a 25-year 25 journey. year uh all in medtech with a little tiny stint in the biotech world. Yeah, I was fortunate to be able to cut my teeth at Medtronic in both uh cardiac and then neuro businesses. And then from there went on to a mid cap urology business that we ended up ultimately selling that was called American Medical Systems. And then a series of startups that really were fantastic for setting me up for what I was able to do here by joining Brainscope, right? And that covered ophthalmology to microbiome therapeutics to some oncology work in

1:48Unidentified speaker (UNKNOWN): a

1:48Matt Adams: whole long list with a couple of nice exits in there and a chance to to join Brainscope just about a year ago.

1:54Unidentified speaker (UNKNOWN): Yeah, it

1:55Chay Carter: was like it was November last year, I think we were saying, right? Um Brainscope itself has been around for a while. It's certainly not a startup straight spot. spinning out of a lab or anything like that. I mean, tell me a bit about the history of Brainscope and what it is you guys are building.

2:08Matt Adams: Yeah, so Brainscope today is a tech enabled AI software platform that focuses on neurocentric biomarker development. EEG is our core measure, although the platform we've built is very multimodal. So we can take in any lab signals or lab data. blood work, for example, speech, imaging, uh genetics, and so on. You know, for those of you that may have known Brain Scope for a while, because as I'll touch on, this is a it's a 20-year-old startup, so it's not brand new.

2:39Unidentified speaker (UNKNOWN): But

2:39Matt Adams: the brain scope of today is absolutely brand new. And in the last year, we've gone through an incredible transformation to frankly go from a single product, EEG. company that had an algorithm focused in detecting brain bleeds to a platform with multiple commercialized products, multiple sales channels, and a pipeline behind this that is leading with dementia.

3:04Chay Carter: That's interesting. So okay, there's a few a few things to unpack there, right? So does the existing brain bleed or brain in is that brain injury concussion? Is that what we would we'd call that? Or what does what does is that still exist as

3:17Unidentified speaker (UNKNOWN): well?

3:17Matt Adams: It does. And so the way the business started was with an FDA cleared algorithm that was designed to detect bleeds in your brain following a traumatic brain injury. That specific algorithm is 99% sensitive all the way down to a single CC of blood. For those that aren't familiar with our approach to assessing that, we have a small point point-of-care of care EEG. EEG array. We have AI embedded on a phone platform. The whole process from opening the package to collecting the signal to throwing everything away takes about 10 minutes. We only need two minutes of actual clean signal from your brain for the AI to run. And it's designed to be done by anybody, frankly. It doesn't need a specialist to acquire it, doesn't need a specialist to interpret it. That algorithm was actually the first FDA-cleared AI-enabled product in the field of neurology. So the business has been a pioneer. That product is still available, and it has got great utility in field applications where you don't have like imaging right around the corner. Rural medicine, same story, where you don't have know, advanced imaging sitting right around the corner. Now, since then, we've expanded quite a bit. So when you ask about the definition of TBI today versus years ago, today it includes both brain bleeds, like we discussed. It also includes the ability to detect

4:37concussion,

4:38Unidentified speaker (UNKNOWN): uh,

4:39Matt Adams: assess severity, and then track someone all the way back to recovery. So if you're an athlete, you're very worried about return to play. If you're an active duty service member, perhaps you're you're retur you're concerned about return to active duty, kids return to school, and so on. So for us, head injury today encompasses both bleed, so very acute healthcare decisions and concussion.

5:05Chay Carter: That's really interesting. So you're able to help people understand the recovery process and how close they are to actually being fully able to return to to work or or or the feeling. field or whatever, whatever the situation might be.

5:17Matt Adams: Uh we can. And so for those that are familiar with that space, you know, there's a number of assessment tools that are available out there, but what you find mostly in use if you get in even to a specialized setting like sports medicine clinics, for example, those tools are typically made up of a list of subjective assessments, right? Where they're doing their best to take a multimodal view of a of a patient and get the the best assessment you can to do you have a concussion and where are you in a recovery cycle and and return to play. What we bring to that clinical assessment category is a functional assessment that's looking directly at brain activity. So you're getting a direct measure of the injured organ

6:01Unidentified speaker (UNKNOWN): and

6:02Matt Adams: where it is in its severity or presence of a concussion, along with how it's recovering as a function of the therapeutic approach you're taking, and ultimately making the decision around are you ready to return to play or return to work?

6:17Unidentified speaker (UNKNOWN): Yeah,

6:17Chay Carter: that's interesting. I I want to ask about the AI side of things because you mentioned obviously the of first algorithm um being approved. Obviously, so much is changing all day, every day in the AI land. landscape at the moment, right? And understanding what it can do with regards to you know signal processing, just understanding the brain in general, really. Data that we've been aggregating for a really long time and now we can actually make sense of it. I mean, what was that? Uh from a business perspective, I appreciate it's pretty before you joined, but I mean that must have been quite an exciting time for them, or pretty quite cutting edge and you know, frontier tech, looking at algorithms for understanding another side of the brain before maybe many others were even you know, contemplating

6:52Unidentified speaker (UNKNOWN): it. Yeah,

6:53Matt Adams: it's a you you nailed that in a if there was ever a company that

6:57Unidentified speaker (UNKNOWN): that

6:58Matt Adams: was born perhaps a little bit ahead of its time and has absolutely been on a frontier at its brain scope. So the business as we reference is about 20 years old. It's not a new startup. But part

7:08Unidentified speaker (UNKNOWN): of the

7:08Matt Adams: reason that early development cycle took so long was point point-of-care of care EEG that could be applied by a generalist. was not a thing. So that had to be developed from scratch. So relatively basic forms, compared to where we are today, that was barely a thing. So all of that had to be developed from absolutely nothing. And then anybody in the AI world today knows that data quality is like always top. of mind. And this is especially so if you're trying to use it to guide clinical decisions. You don't want to be wrong. And so the clinical research cycle to collect high-quality data and make sure that that is building a high-quality algorithm that clinicians and coaches and athletic trainers in the case of concussion can have confidence in. That took a long time. Now, the bonus is. you've got a fantastic platform that you can build from now. The the downside, it took 20 years to get here.

8:10Chay Carter: Yeah. Yeah. It's it's so interesting. I mean how rapidly uh these technologies are expanding too. So it's always fascinating to think like what else you guys can do with this technology because you've got a head start, right? And you've got thousands of hours of patient data and recordings, uh, and you know, all the all of that data you're aggregating. what how what comes in the future for you guys is going to be really interesting. I mean, we we I want to jump into dementia in a second, but I know that you've just returned from London too and had some really amazing meetings directly related to the sport side of of things as well. I mean, tell us about that. How was your trip to London and what can you tell us? I know you just got back.

8:46Matt Adams: Yeah, it's Jay, it was an amazing trip. And so we were over there really for two reasons. In And both of

8:50Unidentified speaker (UNKNOWN): them

8:50Matt Adams: we were invited to come over and participate in a couple different venues. The first was an organization called the Sports Doctors Network. This was founded by the chief medical officer of Real Madrid. Thankfully, we've become close through this process, and it was an opportunity to come and participate with the Sports Doctors Network and be able to speak to um athletic trainers and sports medicine physicians and physiotherapists and so on. from across a number of different sports and be able to bring an objective measure built on a platform like Brain Scope to what is typically a pretty well-rounded list of clinical assessments and a frontier in both functional assessments like a brain scope and molecular assessments, where you might find blood or saliva or other things. coming along as well. We're engaging with Real Madrid now as a function of that invite. Had some pretty powerful panel discussions as well with researchers and other sports

9:50across the pond and across multiple industries. So really exciting there. I would say the interest in being able to bring technologies like we are into collegiate and professional sports that

10:02Unidentified speaker (UNKNOWN): are

10:02Matt Adams: focused. on athlete safety, being able to return athletes to play faster

10:09Unidentified speaker (UNKNOWN): and

10:10Matt Adams: allow for longer careers, like that appetite's really growing. And that showed up in the the second reason we were uh invited to come to the UK, attend the conference, and then actually go up to one of the elite training facilities for mixed martial arts fighters. in the northwest of England called Vortex Performance Centers. They uh have a number of the elite ranked professional fighters that are uh trained in that gym. And of course, they're interested in the same thing. How do we keep our athletes safe? How do we allow them to get back in the ring sooner following an injury? How do we help them extend those careers and live a high quality life afterwards? Which, you know, f for those that know the space, multiple head injuries certainly increases your risk of dementia. And that experience, the world's changed a bunch, Jay, because we we had fighters lining up asking if they could be assessed. And the bonus there was we were able to identify a few that were definitely showing a concussion, and we're able to connect them with a therapist immediately.

11:15and get them into treatment so they're able to recover more quickly and get back in the ring. Perfect real world use case.

11:23Chay Carter: That's incredible. Yeah. I was just gonna say it's um it's so interesting when we talk about neurotech businesses that are growing and you know that we're in this interesting period of time where I think brain health is it's becoming more, you know, more of a logical and obvious and easy conversation to have. But I'm pretty probably sad. saying that as somebody that works with a neurotech similar to yourself, so we talk about it all day, every day. The reality is the general public probably aren't thinking about it. So what's the best way to gain greater access is just literally turning up at these places and saying, Hey, we're gonna do brain scans. And the fact that these people are queuing up for it as well is is is so important. And what's particularly interesting I think is, you know, I grew up playing soccer, well, football where I'm from but well, you know, and no one ever spoke about concussion or maybe any any sort of brain injury because that was always associated with rugby or American football or ice hockey or you know, even you know, cricket wearing a helmet because a ball might hit you in the head, a little baseball, right? It's never associated with soccer. The balls are a bit softer, it's a bit more you know, played with the feet. But over the last few years I've had more and

12:23more interviews with retired professionals talking about issues of. they've had or everyone knows someone who's had an issue from maybe it's a defender who's heading the ball all the time or a goalkeeper who's getting barged into all the time and things like that. And it's just really interesting to see like there is you know, our pretty our acceptance and our understanding of brain health has changed a lot over the years where even sports that we thought were relatively safe are super important to get people, you know, uh checked out on. So amazing to see that Real Madrid too, like the pretty the leading global brand in football is particularly interested in using it. And just it's exciting to think what that uh commercial application means like for you guys.

13:01Matt Adams: Well, for the business, it's a fantastic thing. And for the number of people we can have a positive impact on.

13:06Unidentified speaker (UNKNOWN): Yeah,

13:07Matt Adams: just if you look at youth sports in the United States as an example where everybody clearly starts start before you get to the more elite professional ranks. And there's 12 million youth athletes that are in concussion-oriented sports, right? Where they're at risk for something like this. And to your point, more often than not in those settings, if there's anybody on the sideline of the field or the pitch at all who's who's thinking about head injury, they're likely haven't had the opportunity to get any real education in and concussion, likely don't have a tool to do any sort of an assessment. and often don't have anywhere to send anybody, even if they do find something. And so one of the ways you develop this space is a lot of education. And in the youth sports area specifically, we are we are this far away from launching a program called Every Brain in the Game, which is an opportunity for philanthropic organizations, corporate sponsorships and so on to be able to go to environments like that. So the youth sports environment with a particular focus on the three million youth athletes that are in underserved communities, and be able to provide education, be able to provide a tool that is very

14:15simply and objectively assesses whether you have an injury or not, and then connect those programs back to a sports medicine clinic where they can get care for those that need it. One of the bonuses we have today is we know. that if you can get people into care quickly, this is why there's so much building traction in the professional sports world.

14:33Unidentified speaker (UNKNOWN): This

14:33Matt Adams: is new, right? In the last few years, it was if you have an injury, you kind of go sit in a dark room and wait it out, right? Today, that's not true. If we can identify you early and get you to care quickly, your recovery times can improve by 50 to 73%, depending on who. you read. But dramatic changes. And now that will now start in the youth sports world with our every brain of the game program.

14:59Chay Carter: Yeah. It's like the best ideas are always the most obvious ones, right? Um, and like your youth sport brain monitoring for concussion and even TBI, right? Is uh just so glaringly obvious. We just associate it with professional sports or, you know, uh as adults. That's So that's an incredible initiative. Um what does a treatment sort of look like then post you know post identification of an issue, percussion or brain injury? Like what happens then once you guys spot it? What what's the typical experience for a patient after you know spotting it for you really, really quickly?

15:29Matt Adams: Yeah, great question, Shay. You

15:30Unidentified speaker (UNKNOWN): know,

15:30Matt Adams: it's very much an evolving space because if you think about classic therapeutics, drugs and devices and so on that can go after a condition like this, there's a lot in development today, way more than there used to be. Not that many that have actually gotten to market yet. So for the most part, when you're injured, it's going to be targeted therapy, physical therapy, speech therapy, cognitive therapy, things like that, ocular therapy, depending on what your specific symptom profile looks like.

15:56Unidentified speaker (UNKNOWN): And

15:56Matt Adams: getting you in the hands of a of a professional therapist quickly will get. you recovered a lot faster.

16:04Chay Carter: Yeah. That's awesome. I'm going to talk about the dementia stuff. Um, because that sounds like a an expansion outside of the original plans for the business. I mean, tell me about that. Is it neurodegenerative is is pretty the hottest area, I think, within Neurotech, because it's so chronically underserved by you know pharmaceutical options out there and Neurotech is just about starting to get some traction there. Tell me about the work you guys are doing in that space and and what what advancements you've you've generated.

16:31Matt Adams: Yeah, I mean the neurodegenerative space and specifically the dementia space as you reference, such a massive problem for most countries around the globe, right? I mean today in the US you got 7.2 million that are diagnosed. That number is going to be gonna two to two and a half times that by mid-century.

16:48Unidentified speaker (UNKNOWN): The

16:48Matt Adams: number at risk

16:50Unidentified speaker (UNKNOWN): is a

16:50Matt Adams: about 22 million and growing at the same rate. And so we've uh invested heavily in developing biomarkers that will help predict the onset of disease, assess severity over time, and then being able to subtype different diseases. And so touch on each of those briefly. The prediction biomarker, Marker. we actually have the algorithm developed. It was published earlier this year in Nature Scientific Reports. That algorithm will predict the onset of dementia up to seven years earlier than standard of care. So we're giving clinicians, patients, and clinical researchers an incredible amount of insight and control that they would have never had before. We'll follow that up with the ability to assess changes in severity over time, like we already do in FDA cleared conclusions. concussion products. We'll have this similar model happening in dementia. And then the last component is subtyping. So anybody familiar with the dementia space knows that's actually a collection of different diseases, predominantly Alzheimer's disease, which most of us, if it's not in

17:55our families directly, we certainly have been touched. We know someone who's been impacted by that disease.

18:02Unidentified speaker (UNKNOWN): To be able

18:03Matt Adams: to separate the various forms of dementia, think Alzheimer's from like Lewy body or frontal temporal lobe dementia or vascular dementia, is hard to do clinically.

18:12Unidentified speaker (UNKNOWN): And

18:13Matt Adams: so where you can provide an objective measure of what's actually going on. Again, you've provided clinicians

18:20Unidentified speaker (UNKNOWN): with

18:20Matt Adams: insight that is it's very difficult to get today, and you've provided clinical researchers with tools that will really go a long way towards extending. accelerating new therapy development.

18:29Chay Carter: I mean, that's incredible, the seven year head start over a competition. Not that it's about competition, it's about the patient population, but it's just an amazing advancement in the space. And as you said, everyone, everyone has been affected by it. You've either seen it with your own parents or grandparents or great grandparents, or you know someone that has. I mean, you all know how difficult. it is to deal with. And there's lots of talk about yeah, how do we spot it earlier? You

18:54Unidentified speaker (UNKNOWN): know,

18:54Chay Carter: how do we slow it down? So once we've actually spotted it. There's obviously lots of work being done around the neural plasticity space at the moment, but nothing is 100% foolproof. And things like this are really, really useful because

19:05Unidentified speaker (UNKNOWN): not

19:05Chay Carter: everyone's going to have access to the latest tools and technology to wear at home. And you know, some of those those therapies are really, really expensive. This is this is uh you know, a bit more of like an easily accessible. scan, I suppose. I mean that's that's really fascinating. What's what's the what's the next steps for you guys in this space? So what does the future look like for your dementia treatment specifically?

19:22Unidentified speaker (UNKNOWN): Well

19:22Matt Adams: in our dementia portfolio, so for the prediction biomarker that I mentioned, algorithms develop. We've already validated against two external data sources, so we're teeing up now actually for what we anticipate to be the last validation before we go for regulatory clearance and starting with the US and going beyond. The other two that I reference, both measuring changes in severity and then subtyping, are just out of feasibility stage and moving into formal you know model development and training, kind of typical AI uh model development process uh before they go into validation, which is essentially structured like a blinded clinical trial before we go to submission. So those are both on a very good or all three of those, I should say, on a very good trajectory, with the prediction biomarker being relatively close to market, actually.

20:09Unidentified speaker (UNKNOWN): Oh,

20:09Chay Carter: fantastic. How how close are we talking?

20:12Unidentified speaker (UNKNOWN): Of

20:12Matt Adams: course, there's the normal caveats around development and all that kind of thing, but uh, you can figure we're about 15 months or so out.

20:19Chay Carter: Matt, fantastic. So, what's the like from development to customer, consumer, patient pathway look like for you guys? Because I think obviously I speak to a lot of people in the implantable device space, the wearable device market, consumer device, which is obviously a completely different ballgame altogether, but like the regulatory and clinical pathway for you, what's what does that sort of look like for a for a company doing work like this?

20:42Matt Adams: Yeah, so great question. The vast majority of your time is spent in some sort of data collection. So for us, like, and this is partly why the business was really a long startup process, because gathering all that clinical data was done from scratch. And so you're building a large case of normals, you're building a large case of injured patients, and then you're going through model development, you've got to train it, you've got to then take it through a series of validations, the last of which looks very much like a blinded clinical trial for those that are used to operating in the implantable space for example, before going to the FDA or whatever regulatory body for clearance. Our original product was cleared through a de novo pathway with in the uh FDA. And then our follow-on, of course, were all 510K, so class two

21:33Chay Carter: uh

21:33Matt Adams: products. And that's that included both hardware initially as well as algorithms as

21:39Unidentified speaker (UNKNOWN): more

21:39Matt Adams: and more algorithms have come out. We're now up to eight FDA clearances.

21:43Unidentified speaker (UNKNOWN): We're

21:44Matt Adams: also included in Saudi Arabia. We're including Canada. We are poised to begin a pilot program in China. So the the pull is uh increasing.

21:53Chay Carter: Yeah, that's fantastic. Um you mentioned obviously the hardware and the algorithmic side is too uh brainscope brain scope themselves and uh you're you're a hardware and software provider. or are you using yeah you had your own patented hardware uh or is it just on the um the AI side? Like what does it look like as a as a portfolio?

22:10Matt Adams: So we have the reason I opened with we're a tech enabled AI software platform is the tech is the primarily the sensor. Looks a bit like a sweatband, right? Very, very easy to put on. But at the time that was originally developed, that was very novel technology. And so we continue to produce hardware. We have our second generation sensor in development today. It's bring

22:32Unidentified speaker (UNKNOWN): some

22:32Matt Adams: cost out. It's reusable. It makes it uh even simpler to uh for the user to engage with. It enables a subscription model, so it brings your cost per test down a good bit and creates market access that's pretty unprecedented for something like this.

22:48Unidentified speaker (UNKNOWN): The

22:49Matt Adams: core of the business, though, is really the the algorithm development. You know, the reason that we sell the hardware is because you don't walk into just any locker room or emergency department or sports medicine clinic and find the ability to collect brain waves in the hands of mere mortals, right? Most of us The bus without specialty training, and do it easily. So we continue to sell the hardware. frankly, just to make signal acquisition as simple as we can and make it possible to do in any location.

23:22Unidentified speaker (UNKNOWN): So

23:22Matt Adams: if you're sitting on the side of the football field or the soccer pitch,

23:26Unidentified speaker (UNKNOWN): or

23:26Matt Adams: you're sitting in the clinic or the emergency department, all those environments are possible. That wouldn't happen without the type of hardware that we've designed and produce. produced.

23:36Unidentified speaker (UNKNOWN): Now,

23:36Matt Adams: as the business grows and the data set. sets continue to get larger and larger, which we're well down the road of a substantial multi-disease state data set, our real value comes in the algorithm development and the ability to provide that high clinical utility decision support.

23:57Chay Carter: So it's really interesting because um I think when we first spoke actually one of the first chats we had was about data aggregation, data collection and how that industry or how those words I suppose are becoming way more synonymous just with a regular neurotechnology device, um, as we're starting to understand the brain more and and what's happening in that market. And I think companies historically that have been working on even neurostimulation for the brain are now starting to call themselves or even even diagnostic to they're calling themselves neuro neuroAI AI companies now. And there's like a few more buzzwords picking up and I think neuroAI is an incredibly attractive word to probably a non-traditional med tech or neurotech investor, because it sounds new, modern, and and probably relevant to some stuff that some of the billionaires are investing in at the moment. Um, one of those main areas, I suppose, is like foundational models, seems to be the the new one. You guys are aggregating a ton of data, probably more more so over the years, and than a lot of people who are actually building foundational models. from scratch, really, maybe out of spinning out of labs, for instance. Um what what are your thoughts on that market? Um I mean would you would you ever call yourselves a foundational model for dementia or or brain injury or TBI, for instance?

25:03Matt Adams: Yeah, well, so

25:06Unidentified speaker (UNKNOWN): this

25:06Matt Adams: will be a hot debate as you can imagine, as you as yeah as you as you knew when you asked the question. So there's there's a couple of different ways to think about this. Do we have a foundational model? We actually have a large scale multi-disease state model that works today. You can put in your EEG and get assessed against a whole panel of different disease states, right? That go from head injury like we're already in to dementia that we're entering to neurodegenerative disease like Parkinson's to psych, major depressive disorder, generalized anxiety, ADHD, and a long list below that. So, from a sort of classic foundational model perspective, yes, we have that. Now, what we've delivered historically and will continue to deliver are very specific models that are answering very specific questions. One of the reasons we do that is there's an intense focus on data quality and making sure that the decisions you're facilitating are done. with a high level of sensitivity and specificity. And that drives data quality requirements. That if all you're doing is amassing sort of open data sources, you need to be very careful with, our view.

26:18Unidentified speaker (UNKNOWN): Which

26:18Matt Adams: is part of the reason why we maintain a clinical research capability in the ability to go out and collect our own data that's geared towards answering very specific. questions in a real life clinical scenario.

26:32Chay Carter: Yeah, that's uh

26:34Unidentified speaker (UNKNOWN): I mean

26:35Chay Carter: it's a really amazing way to look at it. I think a lot of the companies I speak to who are building these foundational models aren't looking at it, look aren't looking at it or applying it to a real world scenario yet. They're in the aggregation space, they're bringing it in, they're trialing it on a few things, but you guys are building, have built it. And again, I always think about this with companies who have been. cut aggregating data for a very specific purpose for a long time before the buzzwords of foundational models or neuro AI existed. Um they were doing it to solve a problem. And often it's like now going, okay. It's like when I think about um cochlear implant companies going, Well, we were the first BCI. Um, okay, and they can now call yourself a BCI company comment if you want, mate. But it's fine, you know, there's it's like now go, good. do we need to backtrack and start calling it the new latest? buzzwords? It doesn't really matter. At the end of the day, like you're solving a really important problem, and you have a foundational model for people suffering from concussion, from brain injury. Now you know it'll turn into dementia soon as well, right? I think some so I imagine some neurological cases are pretty easier to spot than others. Um, but the fact that you can spot it seven years before shows that there's a tremendous amount of relevant data there um that has been built up over a long period of time. I mean it's uh yeah, it doesn't we don't really need to call it anything, do we, really? As long as it solves a problem that you guys are setting out to do, and it sounds like you guys are doing that across you know multiple neurological biomarkers. I mean, what what does the future hold then for the the biomarkers that you're looking at right now?

27:56Unidentified speaker (UNKNOWN): Uh

27:56Matt Adams: so if we're speaking about concussion as an example, you know, we'll continue expansion of that into uh well, reaches where we're not. right? So today we're moving into sports medicine clinics and expanding that. And then where there's great utility is if you walked into sort of the everyday American emergency department today, you're most likely to get a CT for a concussion, which isn't going to show you anything at all. And unfortunately, most physicians didn't get the opportunity to get concussion-specific concussion specific training. So we're we focus a lot on providing education in a basic tool in the form of brain scope that any of us can pick up, get a good solid measurement, understand whether there's someone who has an injury or not, and then we embed them in a referral channel. So to get them back to the specialist for care quickly. We'll continue that expansion. We're in NCAA Sports today, plenty of room to expand there. We touched on the expansion in into professional sports just a minute ago. So there's actually more pull there than we were anticipating at this stage of our maturity. And then the Every Brain in the Game program is all about moving into youth sports and starting, especially with a focus on the underserved. Uh dementia will launch into clinics, uh similar to what we're doing with sports medicine, except now you're thinking about memory care and the uh neurologist's office

29:16that's seeing those patients, same referral-based approach,

29:18Unidentified speaker (UNKNOWN): as

29:19Matt Adams: well as clinical trial markets, because sort of the foundation for the business or why we focused in this area is that the subjective assessments touch almost every neurological disease. And too often that translates to you either don't diagnose, you miss one. It translates to either patients never getting treated or getting treated very late. It translates to 90% clinical trial failure rates because these trials are so hard to run. I mean, not only is the therapy challenging, but the measures you're using are difficult.

29:53Unidentified speaker (UNKNOWN): And

29:53Matt Adams: the more we can clean all that up

29:55Unidentified speaker (UNKNOWN): and

29:56Matt Adams: provide simple objective measures, the more faster we improve care and accelerate therapy development.

30:01Chay Carter: Yeah. It's uh it's really, really exciting. I I I particularly think about the the dementia work you're doing and and getting that into more people's hands. I I've been thinking more recently about all these neurotech companies around the world or you know, neuroscience driven businesses and and the the part they play in like a in like a workflow, really. So the earlier imagine spotting dementia seven years earlier than other people were spotting it. And then there's all these neuroplasticity neurotechnology devices being built. I've I've interviewed some people in the podcast working on the sort of TMS treatment for memory disorders. And but it's usually for people who are far down the line. And it's like, how do we slow it down and ideally in the future try and reverse the effects of of Alzheimer's or just just germinalist generalist memory disorders? And like, well, what if we're spotting it seven years before? I always think like there's a workflow and there's a rehabilitation piece. that comes afterwards as well. And there's obviously neurology clinics and so forth. But yeah, it's a lot of that same stuff is happening in stroke patients too. Um there's also, you know, more generalist stuff as well, things like ADHD and and depression, which are starting to get built out. Are are those sort of biomarkers on the radar at all for you guys outside of, you know, Now I said, is that yeah, concussion and Alzheimer's and and dementia and stuff?

31:12Matt Adams: The short answer is yes. So we're already published in stroke,

31:17Unidentified speaker (UNKNOWN): and

31:17Matt Adams: that goes after kind of like you'll see our theme across our platform a very specific clinical question that is foreshadowing the typical build out of biomarkers we have for each disease. So think about that framework as we're either predicting or diagnosing, we're assessing a change in severity over time, or we're subtyping.

31:38Unidentified speaker (UNKNOWN): And the

31:38Matt Adams: In stroke. environment, about 40% of the patients that walk through the door of your emergency department, as an example, with stroke type symptoms, will be having something that's not a stroke. It's called stroke mimic. Those can often be made up of various forms of severe migraine or a syncope, something like that. Of course, in the ED environment, you're going to engage in a whole stroke protocol and things got to move fast, right? So that that first publication demonstrated our ability to separate stroke mimic from stroke patients. So patient care pathways now have much better insight earlier. That will ultimately be followed by the ability to track rehab over time and get people back to normal function, kind of our normal process. We already have data accumulated in major depressive disorder, generalized anxiety, some. ADHD. We've even begun pulling in some multiple sclerosis data. So there's a long list of diseases that sit in our pipeline that will continue to mature as time

32:39moves on.

32:40Chay Carter: Fantastic. Yeah, really exciting. There's so much potential, really, isn't there? And what about other companies maybe using the data to build sort of diagnostics around your platform? Are that something you guys look at or is this super patented and this is your world to build around?

32:57Matt Adams: Uh oh man, I so love you asked that question. So you step back from you know all all we've talked about right now are the algorithms that we've developed on our own proprietary data. And that will continue in the future and we maintain a very strong patent portfolio around the process, some of the specific diseases and things like that. That will continue. But the Brainscope platform, and this is part of the for those that knew the business 20 years ago, and you look at Brainscope today, I mean it is a dramatically different business. This is one of the key reasons.

33:28Unidentified speaker (UNKNOWN): And

33:28Matt Adams: it's that platform is really built to be like the operating system for others that want to be able to come in, use our infrastructure to be able to ultimately develop their own. biomarkers, and being able to now to deliver it through phone-based app and sensors like we have, our deep learning model, which is the large-scale, like 170,000 patients across multiple diseases. I referenced this earlier. That's in development. It runs, it works today. You can attach that to any of the Frontier AI. We are in our demo mode, we're currently attached to cloud. Claude. But that means you can now engage from nearly anywhere with your own application that pulls

34:11Unidentified speaker (UNKNOWN): on

34:11Matt Adams: our model sitting behind, right? Using our infrastructure. So long answer to our own specific algorithms will continue to be developed and can continue to be delivered across multiple channels. We also have for our building, the operator. operating system for others to be able to come in and use all that infrastructure to develop their own.

34:31Unidentified speaker (UNKNOWN): One

34:32Matt Adams: of the reasons why we also pay attention to staying very multimodal. We're EEG centric, obviously, as we've discussed.

34:39Unidentified speaker (UNKNOWN): But

34:39Matt Adams: even our dementia work today, I mean, we we pull in blood work. So for those that follow that space, the investment into blood biomarkers and dementia specifically has been incredibly. incredible. And they're making great progress, being able to get more and more specific around molecular pathologies that indicate higher risk for dementia down the road. They don't touch on the functional assessments, right? Which is where a brain scope comes in. You bring those two things together in a lot of environments, you have an even better assessment of what's going on with that patient. Are platforms designed to accommodate

35:17Unidentified speaker (UNKNOWN): that? I

35:19Chay Carter: think this is what I love about what you're building because it's it's still very modern take and right cutting-edge technology now, even though it's, you know,

35:27Unidentified speaker (UNKNOWN): way

35:27Chay Carter: ahead of its time when it was built years ago, it still feels very cutting edge to what you're trying to do and build and uh and And of a still maintains a modern approach to it, but it's also still very patient-forward. patient forward. Um, you're still trying to solve really complex problems such with that we are a long way away from solving, but this is the way you're going to solve them by having, first of all, large data sets, analyzing them using modern science modern uh um software techniques, but also at the same time allowing it to be sort of open source to some degree, uh and allowing people to access that information and plug into it too, to develop their own, because at the end of the day we can't. yes, it's a commercial world that we live in, but we can't forget what we're trying to do here is you know spot these things and makes people's lives better. It's really fascinating. I really appreciate it. And I'm conscious about time, Matt. And one more question for you.

36:16Unidentified speaker (UNKNOWN): If

36:16Chay Carter: we were having this conversation, say in five years' time, what do you think we'll first of all, what do you think brain scope is going to look like? And also, what do you think our understanding of the brain born board looked like? Things that we can't maybe measure today. the way that things are going, the trajectory that you've seen yourself in the last 25 years, what do you think the next five years look

36:33Unidentified speaker (UNKNOWN): like?

36:34Matt Adams: You know, the the great news uh relative to that question is that we know where we're going. So, this is about how fast can you build the right tools to get to that endpoint? And I think the simplest way we think about that endpoint is look at the cardiac world, right? To have a clinician. would completely expect to be able to have their clinical assessments, their functional assessments, EKG. Now, the bonus there is you don't need AI to decode your EKG. You can do that visually. And then a whole long list of molecular assessments. And they're bringing those things together to get a complete picture of the patient and develop a care plan. The neural world, clinical assessments, are really well developed. The functional assessments are just, to your point, just starting to become a thing, right? And that's us. Where because of modern AI, we can decode EEG and give it the same utility

37:31Unidentified speaker (UNKNOWN): that

37:31Matt Adams: EKG's had forever.

37:34Unidentified speaker (UNKNOWN): The

37:35Matt Adams: molecular assessments, there's blood tests out there for their early stage yet, but there's blood tests out there for TBI. There's, as we touched on, for dementia, there's tremendous amount of investment going into trying to define blood-based biomarkers for that disease state. That will continue. So we're going to end up looking very much like how the how the cardiac world has evolved. It's just these tools are not yet as mature, but they're coming.

38:01Unidentified speaker (UNKNOWN): And

38:01Matt Adams: we're going to play a big role in it.

38:03Unidentified speaker (UNKNOWN): it.

38:04Chay Carter: Yeah, I love it. I interviewed uh Ken Nelson, who's doing some work in uh Epitel these days, but he doesn't a strong background in uh the cardiac expression. His message is exactly the same. It's like we need to look at the cardiac market and how we developed that and how long it took and how how broader it got as a model to follow for neurotech. And I often look at brain computer interfaces, for instance, and go, hey, look, neuromodulation is a as good way to follow. that but I think when it comes to data and it comes to brain assessments we've been doing this stuff in other industries for a really really long time and I think it's uh it's an absolutely valid comparison for sure. Um Matt, thank you so much for your time man. Um amazing to see you again and what an amazing story. Thank you.

38:43Matt Adams: Jay thank you very much for the time and always a good chat with you.

38:48Chay Carter: Cheers.

38:49Matt Adams: Cheers

38:50Unidentified speaker (speaker_2): This podcast is brought to you by Carter Sciences, the headhunting practice that covers not only headhunting but also fractional talent work for startups who need help but don't want to spend those big recruitment fees, wanting to work on a more hourly basis with companies and also career coaching for neurotech professionals. Goal is to provide cost-effective solutions across all neurotech hiring. needs.

This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights