To the 47th Annual Raymond James Institutional Investors Conference. I'm Jayson Bedford. I cover the med tech sector here at Raymond James. Thank you all for joining us. I'm really pleased to introduce Scott Blumberg, the CFO of Ceribell. Ceribell's the first-time presenter here at the conference, fast-growing small cap company that we like, and I think you'll enjoy the story. Scott will go through a presentation, and then we'll have a bit of Q&A at the end. With that, it's all yours, Scott.
Thank you, Jayson. What Ceribell does is we've developed a novel point-of-care EEG system targeting neurological conditions in the acute care setting. We've got a form factor that makes EEG much quicker and easier to acquire, and we've overlaid it with Clarity, which is a machine-learning AI-driven algorithm that can detect seizures at the bedside. Just to provide quick grounding on where we are commercially, we just wrapped 2025, in which we delivered $89 million in revenue.
That represents a 36% year-over-year growth rate, and we're operating at high gross margins, 88% last year. Throughout our history, and we started commercialization around 2018, we've largely operated in the U.S. market for seizure detection in the acute care setting, which is the ICU and the Emergency Department. That's roughly a $2 billion total addressable market.
Throughout the course of 2025, we had a number of milestones aimed at expanding our indication, including expanding the algorithm to neonate and pediatric patients, as well as getting the first indication for delirium detection, which represented a $1.5 billion TAM expansion, bringing our total addressable market up to $3.5 billion. For a quick framework, the way we think about our growth trajectory and vision is in three horizons. Our goal right now with our current product and seizure is to become the standard of care for seizure detection in the acute care setting.
In the future, in the near future, as we introduce other algorithms onto our platform, our goal then becomes to make EEG a vital sign, which is making EEG so prevalent for the indications that affect people in the acute care setting such that it's used broadly across anybody who's altered. Over the long term, and I won't get into this too much today, we also have ambitions to go beyond the acute care setting to other sites of care as well, as well as develop brain biomarkers. First, I'm gonna focus initially on the seizure, the seizure market. A lot of investors, when they hear seizure, they immediately think about epilepsy, and that's not exactly where we play.
We play in the acute care setting, the ICU and the ED, in which in which area seizures are not typically caused by epilepsy alone. Epileptic seizures are visible. You typically see the patient convulsing. They're often shorter in duration, lasting seconds or minutes. Although it's a very, very serious condition, they're not-- A single seizure is rarely life-threatening. The exact opposite is true in the acute care setting on all dimensions. First, the seizures are non-convulsive, meaning that you don't display the hallmark symptoms that you typically associate with seizure. They can last much longer, hours or days even. Because of how aggressive they are, they can be life-threatening to the patient. They're common across a wide number of conditions.
The list of conditions up here that create risk for seizure is broad. It includes conditions that you would often consider neurological, like seizure, like prior seizure, brain tumors, stroke, traumatic brain injury, but also a number of indications that aren't necessarily typically considered neurological, like cardiac arrest or sepsis. The commonality amongst all the conditions is whenever there's an acute assault on the brain, seizure is the response. In 92% of these patients, the seizures are non-convulsive, meaning they're typically not visible with the naked eye empirically. For these patients, EEG is the only means to diagnose the patients. Speed of diagnosis is critically important in this market. The length of a patient seizing, and we use the term status epilepticus, which is a society-defined term meaning a seizure lasting 5- minutes or more.
A patient lingering in status epilepticus or prolonged seizure can have a significant negative impact on the patient's outcome, both in terms of mortality and morbidity, meaning permanent loss of memory, loss of IQ, loss of cognitive function. The need for prompt diagnosis is very important. To the same end, the quicker you administer the first-line treatment for seizure, the more effective it is. Initially, the treatment is readily available. It's a large dose of Ativan, typically. If you delay even 1 hour, the treatment effectiveness goes down, and it goes down further the longer you delay. That results in a significant need for rapid diagnosis detection, and treatment. The society guidelines have come to terms with and appreciate this need.
In 2012, the Neurocritical Care Society suggested that when status epilepticus is suspected, seizures should occur within diagnosis should occur within 15 - 60 minutes. AHA followed suit in 2020, and American Stroke Association in 2021 and 2022, all recognizing the need for immediate detection using EEG. Conventional EEG was not designed for this purpose. It was designed for outpatient diagnosis of epilepsy and we think misappropriated for use in the acute care setting. It's not set up for speed. Conventional EEG requires specialized EEG technicians to set up and administer the technology. They're typically only staffed during limited hours, often nine to five within the hospital.
When they get to the bedside, which can be hours delayed, there's a very lengthy setup process, typically 30 minutes plus to set up an EEG. That's just signal acquisition. You also then need interpretation. Once the signal's acquired, you need a specially trained neurologist or epileptologist to interpret the signal, and they're often just looking at it once. They're not continuously monitoring the patient. What you get is the polar opposite of what patients need, what the society suggests. You get these lengthy delays that can last hours. If you happen to have a seizure in the evening on a Friday at a community hospital, it can be days, and that is what we're fixing.
We have taken that lengthy, cumbersome, and bottleneck-filled process, turning it into about a 5-minute process, rapid signal acquisition, early detection, and continuous monitoring. This is the system. The system's composed of a single-patient disposable headband, and we just introduced a head cap for babies. It plugs into the recorder on the bottom. That's the capital that's reused. Together, they acquire the signal, and they port that data through the hospital's Wi-Fi to the cloud to a portal where the raw EEG data is accessible by the neurologist on any device. This solves the first bottleneck. This is signal acquisition, and it allows for that within 5- minutes versus what can be hours of delays. For interpretation, we've also introduced Clarity, which is our seizure detection algorithm.
It overlays both the physical recorder as well as the portal and allows for interpretation of EEG by non-specialized clinicians at the bedside. I'm gonna dive a little deeper into Clarity. What you see on the top left here is that's raw EEG data. That's what neurologists sit and read. This screen is about 15- seconds of data. When a neurologist reads an EEG that's, call it three, four, 12- hours long, they're clicking hundreds of times through screens. We convert that data, which requires years of training, into seizure burden trend, which is the line on the bottom. I would infer that everybody in the room can understand what that means. When it's high, there's a lot of seizure. When it's low, there's no seizure.
What it means technically is the amount of time in the past five minutes the patient's been in seizure. Red above 90% is consistent with status epilepticus or a seizure lasting four and a half out of five minutes. When status occurs, not only at the bedside is the clinician shown this waveform, but also an alert goes off, you can see that in the top right in the red, which allows the nurses and clinicians to respond very quickly. For a more tangible example, here's a real patient case. First, I will point out on the bottom left that this started at 1:00 A.M. In the significant majority of hospitals in the U.S., there would be no access to EEG at 1:00 A.M. They just don't have the technicians staffed. That alone, saved the patient hours of waiting.
When the device was put on the patient, it immediately shot up to 100% seizure burden, meaning the patient had been in continuous seizure for the entirety of the 5- minutes the device had been on at that point. With this information, the bedside team was able to treat promptly. You can see the pink line there. That's the treatment annotation within the device. It didn't resolve the issue. If that had happened with conventional EEG, the neurologist would have looked at it, hours later would have suggested therapy. No one would have known most likely that the seizure had not been resolved. The device alerted again 30- minutes later. Treatment was escalated with the same drug. Over the next couple of hours, the seizure was resolved.
After resolution, the seizure returned at about 4:30 A.M. They switched drugs and then permanently resolved the issue. Without Ceribell, this patient would have certainly been in seizure, for much longer, very likely throughout the whole night, and as you've seen on the prior slides, that can result in significantly negative outcomes for that patient. We just got clearance for our device for patients below the age of 18. The physical form factor was available for pediatrics, but the algorithm was not, and now we have the algorithm available for children. We also have a specialized head cap for infants. This is around a $400 million market expansion opportunity.
We have initially piloted it with some sites within our install base, which is around 647 hospitals in the total install base, not the pilot. We are prepping for a launch, which we're planning to initiate in Q2 this year. The needs in the neonatal population are very similar, but perhaps more profound than the adult population. Seizures are still common. Seizures are still primarily more than 90%, non-convulsive. Physicians are actually worse at empirically diagnosing seizure within the neonatal call point because they don't have the baseline. They don't know what normal is for a preterm baby. For this reason, the society guidelines are even stronger.
They suggest that even without clinical evidence of seizure, if, even without clinical evidence of seizure directly, if the patient has a condition that causes risk of seizure, you should use EEG. The reason for that is the impact on the patient. What you see here is that 1- hour of seizure can lead to significant and potentially permanent neurocognitive impairment. This shows the developmental delay threshold for language. That's also true for cognition, and it gets worse and worse the longer the patient's seizing, so the need is very significant. I'm gonna move to our commercial model. We run two revenue streams. One is our subscription, which is a fixed monthly subscription for use of the physical hardware as well as the software and the AI.
Roughly 25% of our business that operates at about a 97% gross margin. The remaining 75% of our business is the single-patient disposable headband, and that's typically used for one patient, can be taken on and off of that patient, but thrown away between that patient and the next. The revenue curve and predictability of business is really strong. I think that's hopefully self-evident in the slide up here. We've had 31 quarters commercially. We've had 31 quarters of sequential growth. That's a direct reflection of the revenue model itself. We've got consistency in the SaaS portion of the revenue. It's just fixed. Unless a customer cancels, it just comes in every month. The reorder patterns are very strong and consistent.
It's also a reflection of our high customer retention rates. Despite our success, we're roughly only 3% into this U.S. seizure market. The way we get there is we are in about 647 hospitals right now. There's roughly 6,000 hospitals in the U.S. that have either an ICU or an Emergency Department, we've got a long road ahead of us there. Within our existing accounts, within those 647 accounts, we're roughly only at 30% penetrated. We've looked at that a number of ways, and one way is comparing our top users to average users when controlling for the size of the hospital. We see across all size cohorts that our top users routinely use 3X.
We view that as an upside opportunity to keep driving deeper within our accounts. That's the core of our commercial mission here. It's to get more accounts. We invested in an expansion of our sales team last year. We took the territory managers up from roughly 35 territories in mid 2024 to 55 by Q3 of last year. Those reps are starting to become productive this year. We've also got the FedRAMP clearance for sale within the VA system, which provides another lever, and we've got the playbook refined, it's really a matter of continuing to execute. On driving usage, we really do that in three mechanisms. First is training more docs, training more nurses, day shift, night shift. Second is expanding to other departments within the hospital.
We are in the ICU in majority of our hospitals. We can expand to other specialized ICUs, the step-down unit, the Emergency Department. Third is building awareness and protocols for patients in new patient populations where there's risk for seizure. I'm gonna use the remaining five minutes or so to talk about the second horizon here, which is making EEG a new vital sign. When a patient comes into the ICU, one of many things can be causing neurological malfunction. They could have a stroke, they could have seizure, they could have delirium, it could be something else. All these conditions present very similarly. The patients are altered, confused, slurring words, movement abnormalities, and it's very, very hard to differentiate between these conditions and the treatments. Both the severity and the treatment path are very, very different.
We wanna be able to diagnose and differentiate between all these conditions, change the equation from, "I'm looking for a seizure, I'm going to use Ceribell to see if this patient's seizing," to, "Something's going on with this patient. Let's use Ceribell to figure out what it is." We made significant strides in terms of progress on this dimension last year. We received FDA clearance for delirium in November of last year. We're in process of submitting a New Technology Add-on Payment, which may drive favorable reimbursement for use of that technology. That's about a billion-dollar TAM market expansion opportunity. Then for stroke, we received FDA breakthrough designation, which was announced in January of this year.
In order to get that, which gives the company expedited review and the pathway potentially to additional reimbursement, we had to submit early technical feasibility data to show that there's promise of the algorithm working, also the FDA had to determine that this was a novel life-saving technology, so we cleared both of those bars. I'll just do a deep dive on delirium or maybe a shallow dive on delirium since that's the more near-term opportunity. Delirium's generally often called a acute brain failure. It is organ failure. It's organ failure of the brain. Physicians know how to and treat organ failures of other organs, kidneys and lungs, for example, but the brain often goes overlooked. It's a really common condition.
There's roughly 3 million patients in the U.S. each year who suffer from delirium, that's about 30% of the ICU patients and about 80% of patients who are on mechanical ventilators. The stakes are quite high. Every day you're in delirium in the ICU carries about a 10% increase in mortality rate. If you have delirium during your ICU stay and survive, you have about a 60% chance of developing dementia after leaving the ICU. There currently is no approved device diagnostic for this condition. What's used is called CAM-ICU. It's a survey protocol the nurses administer where they ask the patient a number of questions to try to assess cognitive function. When used perfectly in perfect conditions, it's quite good. The ICU's quite chaotic.
It's not used perfectly. It's only done once or twice a day. It's binary. It just says yes or no, in terms of whether the patient has delirium. It does not talk about the degree of delirium. It's cumbersome. It takes the nurses time out of their busy day. We aim to use the same platform, the same call point, the same commercial infrastructure to drive this market. There's huge overlap between delirium and seizure, both in terms of patient populations, but also presentation. As I mentioned, the patients look quite similar that are seizing and have delirium. The treatment is very, very different. In fact, Ativan, the leading drug for seizure is a delirium-inducing agent. If a patient has delirium, you may wanna think about another therapeutic choice.
About 48% of patients who have seizure in the ICU also experience delirium, and about 42% of delirious patients have some sort of abnormal activity, seizure, a seizure-like activity within the brain. You know, we think this offers an opportunity not just to expand to new patients, the billion-dollar additional TAM, but also to create synergistic value where the sum of the two indications is more than the parts themselves. For this coming year, we are currently engaged in a pilot on delirium. We tend to pilot things for about a year before we launch them. We refine our pricing approach, our commercial messaging, ensure the product is product market fit is excellent, build up the clinical literature. We're in the process of that now.
We plan to fully launch delirium towards the end of this year or in Q1 of next year. For stroke, we got the breakthrough, we still have to run through the clinical and regulatory pathway, we plan to advance that. We're also looking at further advancements for our core platform technology as well, the hardware. Future's very exciting. We're only 3% into this core market, $2 billion. We've proven the need. We've proven that we know how to sell it.
The expansion into pediatrics and neonates increases our seizure TAM to about $2.5 billion. In 2027, when we fully launch that product, we add another $1 billion in with delirium. Beyond that, we've got stroke, we've got other undisclosed indications, this is U.S. alone. At some point, we plan to explore commercialization overseas as well. That's it. Thank you all for your time, and I appreciate it. I'll turn it back to Jayson for questions.
Yep. Thanks, Scott. Well done. maybe just to start, the market opportunity is large, 3% penetration is clearly low. you kinda carry the burden of the innovator here, at least in the rapid EEG space. I guess to someone who's new to the Ceribell story, why isn't adoption higher? if you can just talk about the sources of resistance so far.
Sure. Yeah, we're changing the paradigm of care. EEG has literally been around for 100 years. It was not designed for the ICU setting, but it's been appropriated there. We've got a lot of stakeholders. When we go into a hospital, we need to sell to the ICU doc and the ED doc, who are the ones making the care decisions, the nurses, who are the one who typically put the headband on, the neurologists, who have legacy domain ownership of the EEG, and then the C-suite, who once you build the whole clinical buy-in, the discussion really revolves around cost. It's a bit of a lengthy sales cycle.
The flip side is once it's in, it's in. We don't spend much time at all, doing maintenance. Most of our activities after we launch is on growth. We do effectively zero case coverage. It's a worthwhile investment. It is a process to build that consensus.
When you look at your high utilization accounts, what are they doing different in terms of is it a matter of broad training, increased participation of these constituents? What are they doing differently?
I'd start with that. It starts with advocacy. You need really good buy-in. You need not just somebody who believes in the product, but who's willing to vocally support the product and preach the gospel to their peers. Of course, we can influence that, but in terms of what we can more directly control, good workflow design, making sure that the device is easy to order. That goes to things like implementing an order set so that there's a dropdown that says rapid, you know, point-of-care EEG as an option. Making sure that the bedside clinicians own the budget, so they're empowered to use it without having somebody else sitting over them and saying, you know, "Hey, this is hitting our budget, but you're seeing the benefits." Expansion into all departments.
We see that within our biggest hospitals, we're in both the ICU and the ED, as well as, in some cases, ancillary departments as well. For smaller hospitals, it may be an ICU-only device. Being a frontline solution. In our early evolution, you know, years and years ago, we used to position as a night and weekend solution. That was the path of least resistance, but that is not the key to driving broad adoption. We believe the technology presents as and should be viewed as a frontline solution. The hospitals that really adopt that full belief set and have the right processes set up, those are our best users.
Okay. Okay. Just in terms of the pipeline and the introduction of new parameters, whether it be delirium or new indications such as pediatrics and neonates, how does the introduction of these new parameters kinda aid in the value proposition?
I think it depends on which product you're talking about. NICU, neonate is really just a population expansion. I think there's an argument that says, there's some value in terms of a broader hospital-wide adoption, and, you know, if you're using it here, then you should be using it over here, and there's some cross-pollination of ideas. But that's really more akin to expanding another department within the hospital.
When you talk about the indication add-ons like delirium and in the future, potentially stroke, you're talking about delivering more value through the same platform with to be determined, but certainly only a marginally higher or, or no higher cost. That provides the-- For doing the same thing, they get a whole lot more value, and we think that'll drive growth, not just in our current patient population, but also add new patients into the mix.
I guess you kinda answered it to some extent, but is it more revenue per account, or do you expect these new indications to accelerate new account adds?
It could be both. The We've decided on the pricing model for neonate. One of the things we're assessing in the delirium product is the pricing model itself. For neonate, we charge marginally more, but not double, for the addition of the neonate algorithm onto the platform. And then there's the head cap, which is priced higher than the headband. But we have not made the decision yet on what we do with future indications. To the extent we add more departments, more indications in the same hospitals, that would reflect itself as effectively more revenue per account. It would not reflect itself as more accounts. The way we report accounts is a number of hospitals, no matter how many departments they're using.
Of course, more headbands or more head caps drives product revenue per account. As far as account acquisition goes, there are certainly value to having more technology. Somebody might be favorably inclined to one side of the product, but less favorably inclined to the other. They may be, you know, favorably inclined to both, but it might not be individually enough to get them over the hump and having more will drive it. There's the cross-pollination. If you get into the ICU or get into an ED, it's pretty easy to go down the hall after three or four months and say, "Look what they're doing over there. Look at the value they're seeing. Would you like that value too?
Just to be clear, I imagine, there should be a lot of sales leverage here as you're ramping. It's the same 650 hospitals that can use it for different indications.
Yeah. We're selling, all of our current pipeline products through the same sales channel. We're investing in a small number of folks to aid, and product specialists. They're not the front line, but they can help the reps speak the right lingo, get the marketing message right. Yeah, we get great leverage. We've invested in a very strong, very highly qualified, sales org, and we're putting more products through the existing channel.
Okay. Scott, can you just talk about the competitive landscape and where Ceribell fits in?
Yeah. There's been competitors in our in our market since we were born. A number of startups tried to do things like this back in the 2016, 2017, 2018 timeframe. We're the category creator and market leader. There are a couple of entrants who are active in our space. They've been active for multiple years in one case and over a year in the other case. We see that we're still winning business, that we're not losing accounts, and we feel very confident in our position.
Is it your view that the pipeline here can emerge as a differentiator in those competitive situations?
Absolutely. If you think about our competitive mode, of course, you've got core IP. You've got also the work that goes into building that consensus and getting a program launched at a hospital. Another advantage of the higher lift on the front end of the sales process is that it's very hard to displace. We don't believe you would adopt two systems at once. We think it'd be very hard to displace unless the value is significant.
The biggest contributor probably to our competitive mode is the algorithm. We've got both the expertise, the data and the algorithm, which we think is a big advantage. To your point, Jayson, as we start layering these other indications on, we're not aware of any competitors who are working on broad neuromonitoring platforms that touch the same areas that we do. Offering more than our competitors is another way to stay ahead.
Okay. The gross margins in this business are very attractive. Top line growth is very attractive. Can you just speak to sources of leverage below the gross profit line and ultimate path to profitability?
Yeah, absolutely. The vast majority of our sales infrastructure investment, which is the biggest part of our P&L, is focused on growth. The territory managers just do growth. They focus on first approach all the way through to launch, then they hand the account off. They have zero ongoing management responsibilities. The clinical account managers, although they do manage accounts on an ongoing basis, most of what they do is growth. They expand departments, they find new doctors, they help hospitals implement best practices on protocols. They're doing very little babysitting, very little case coverage. As we think about the future, of course, we've got a very strong capital position of $159 million post IPO.
We are burning, but we're burning a reasonable sum. We keep our eyes towards the ability to control our own destiny, ensuring that we don't run a position where we need to raise capital. That's something we feel very confident in. We calibrate the investment in an infrastructure investment by choice based on the best available data we have at any given time.
Okay. With that, Scott, we're hitting the bottom of the hour. The breakout session will be in Amarante I downstairs. Thanks for all of you for participating.
Thank you all.