Advice Analyst at Bank of America. I wanted to introduce our next speaker, Jane Chao, CEO of Ceribell, a newly public company. Jane, welcome, and go through a slide presentation.
Thank you, Travis. I’ve got this mic. I think it’s good. I don’t need to—if you cannot hear me well, I don’t need to be on the mic, right? Okay, perfect. Good morning, everyone. Thank you for being here. I am very excited to share our story. It’s our first time joining BLB conferences as a public company. I’m going to walk you through the unmet clinical needs and our solution, the business model, commercial success, and briefly on our pipeline. At the highest level, we invented a novel brain monitoring system that captures brain waves, or EEG. It has the hardware that is very easy to set up. Nurses, really anybody on the bedside can set up in a few minutes, which is a huge improvement of the current standard of care, which we will talk later. It also has the algorithm front. We call it Clarity.
It's an an AI-powered seizure detection algorithm. For now, the system focuses on seizure management in acute care ICU and ED. In the U.S. alone, that’s about a $2 billion TAM. We also have a very strong pipeline, mostly involving EEG signals and AI for indications other than seizure. For now, we still focus on acute care, potentially detecting stroke or delirium in ICU and ED, which I’ll discuss later, significantly expanding the TAM. In Q1 2025, our revenue was $20.5 million, a very strong quarter. Active accounts in the U.S. only: 558 hospitals. That translates to 42% year-over-year growth in revenue, with a company-wide gross margin of 88%. Now, switching gears, what’s the unmet need we’re addressing? Seizure is very common—a third of neurological patients in ICU experience seizure.
It also goes beyond neurological patients. If you look at it, the first one is following convulsive status epilepticus. You're going to hear me saying status epilepticus or status a lot. That essentially is seizure lasting for five minutes or longer, long seizure, and is considered as a neuroemergency. Convulsive seizure means patient having symptoms, right? If patient have a convulsive status, everybody drop everything, run it. If patient can have status without any symptom, that's non-convulsive seizure. 50% roughly patients after convulsion would have non-convulsive seizures. Also goes brain tumor, TBI, post-stroke. These are all the neurological patients. Somewhere between 10%-30% of them have seizure. Beyond seizure, sepsis patient, post-cardiac arrest patient. The common theme is when patient have a big assault in the brain, it's autoimmune response. Majority of these patients, when they have seizure in ICU, they have no symptoms.
Up to 92% of them, they are non-convulsive. You have to have an EEG to diagnose. You can't just guess it. We often say time is brain for stroke, and our physician often say this for status. Here's why. If patient sees 10-20 hours, mortality rate can go up as high as 85%. The gray bar is when they survive, it's a secondary brain injury, often irreversible brain damage, memory loss. You might wonder which modern ICU would allow your patient to see for 20 hours or longer all the time in the United States, in the neighborhood hospitals. You don't have to go to rural area. I'll explain to you why later. The good news is, similar to stroke, if you can detect early and treat early, treatment is very effective.
If you treat patient within the first 30 minutes of seizure onset, 80% patient respond. If you just delay that by two hours, half of the patients stop responding. Very similar to stroke, early detection is one of the most important factors for patient management here. That means early EEG because it's the only way you can detect seizure. However, the current standard of care for now is the conventional EEG. This year is the 101-year anniversary for conventional EEG getting invented. Nothing really has changed. It was invented for epilepsy diagnosis in outpatient clinics. It's not invented and designed for acute care setting in ICU and ED where every minute counts. It has three fundamental bottlenecks. A, you need an EEG technician to set up. Nobody can do it. Most hospitals, they have a few of EEG tech Monday to Friday, 9:00 A.M. to 5:00 P.M.
That's 25% of total hours. Seventy-five percent of the hours, that's not Monday to Friday, 9:00 A.M. to 5:00 P.M. Most hospitals have no EEG coverage. It takes a long time to set up. Also, EEG monitoring doesn't help anyone. You have to diagnose. The majority of physicians, even neurologists, don't know how to read it. There's a large delay. Only epileptologists can read it. With that, often getting a conventional EEG can take—it's always taking hours and very common takes days in patient arrival Friday afternoon. That's not what the guideline require, which is about an hour to get rapid EEG. That's what we change. From having seizure suspicion to getting an EEG set up, it's only five minutes. Clarity, our algorithm, would tell right away to the bedside your patient in status or not in status.
It will continuously monitor every 10 seconds, doesn't get tired. This is what the device looks like. This is the hardware. Instead of having EEG tech, nurses, residents, really anybody on the bedside, after a few minutes, 10 minutes training, they can set up EEG in five minutes or shorter. The device would stream the signal through hospital Wi-Fi through our portal, and neurologists can read in real time. Clarity is the new component conventional EEG doesn't have. That's the AI read EEG constantly. That's a lot of information. I'm going to show you a real patient case, see how all these features actually deliver clinical impact. This is a real patient case. The first thing you see on the lower left is 1:00 A.M. Majority of the hospital in the United States, everywhere, cannot get an EEG at 1:00 A.M.
That already made a big difference. Then, within the first 30 minutes, the device turned red and sent alerts saying patient is having long seizures. Again, even if you get an EEG, majority of hospital 1:00 A.M., nobody can read it. You have to wait often the next day. In this case, patient was in status. You see the little pink needle there? That is a real bedside annotation. That is when nursing treated the patient with first-line medication of anti-seizure. It happened within a few minutes. That is just very unlikely to happen. As you recall, patient does not always respond to the first-line medication. In this case, in an hour, device came back, said, "Hey, your patient did not respond. Patient is still in status." Then it took them again a few minutes. They treated the patient.
You can see seizure burden drop in real time and bedside know, "Okay, patient responded this time. Seizure is dynamic. At 4:30, patient returned to seizure." This patient, and again, they treated right away, controlled the seizure. This patient survived the ICU, let's not say, got discharged later. She would have seized 20 hours easily because you won't get an EEG. You get it in the good center. You get an EEG next day. The neurology reads a few hours later if you're at good center. If you're in the under-resourced center, you transfer patient now, it stays out delay, almost game over for this patient. It happens all the time. That's anecdote. Systematically, we have published many evidence. The first row you saw is we constantly reduce over-administration of anti-seizure medication.
Because when doctors know this high mortality, morbidity, they can't get the EEG fast enough, they just empirically treat. And this medication often leads to intubation, ICU admission. We showed from community ED to large teaching center, near ICU, somewhere always 40%-50%, we changed patient management after using Ceribell. That's a big change when you think about it. Half of the time, doctors thought they want to intubate, they don't intubate. Docs thought patient's seizure free, they end up having seizure. And then as a result, we show significant reduction of length of stay. The biggest, strongest publication came out last year, 4.1 days length of stay. And each ICU day can cost hospital $4,000 or $5,000. We also significantly reduced patient transfer. Again, majority of these hospitals, Monday to Friday, 9:00 A.M. to 5:00 P.M. So they don't have an EEG, they transfer patients out.
We reduce majority of them. The clinical evidence and clinical value really translated to our adoption. This is, I'm not going to read the number. You can see the past eight quarters, nine quarters of our performance. A few things I want to emphasize. You can see every single quarter, we have not just rapid growth, but also very steady growth. We do not have a single quarter that's lower than previous quarters. There are a few factors behind it. One is our business model. We have 25% of our revenue is from the SaaS. It is a subscription fee, hospital pay for our algorithm and seizure detection AI, and that have very high reoccurring rate. The other 75% is all disposable. That is when they use the headband that is a single patient use. It is a very steady usage pattern.
That's why we keep seeing the majority of our revenue have a recurring and very consistent growth. That speaks for the strength of our business model. Moving forward, we are very excited about our pipeline. Before we talk about our pipeline, I don't have a slide for it, I want to emphasize that despite all our great growth, we're only 2%-3% in the US seizure detection market, right? Because the total TAM is $2 billion. In the next few years, we're going to continue laser focus on seizure in the US, in acute care. That being said, we also significantly invest in our future pipeline. On seizure management alone, we are continuing more expanding to different younger patient population. We already have hardware that covers all the ages, but Clarity, till a month ago, only covers adult.
Last month, we are very excited to receive the pediatric Clarity expansion that covers age one and older. It is the first and only FDA-cleared seizure detection algorithm, covers such a young population. A very unique thing is this data was validated, not trained, validated based on 1,700 patients. Based on FDA records, the biggest by far patient base that are used for validation of the algorithm and speak for the rigor of our algorithm. We are also working on expanding Clarity to neonate as well. Many preemies are subject to having seizure, and the neurological outcome for these patients are in many ways even more significant. Moving forward, as I mentioned earlier, we are also working on making EEG a new vital sign in ICU. It goes beyond seizure. Potentially detect delirium, potentially triage stroke.
Just like if patient have chest pain, you want to EKG on. If patient have altered mental status, for doctor, they can't even tell. The first thing they want to triage and rule out is stroke. Then it's seizure and status. Then it's altered mental status, encephalopathy, delirium. The EEG is the richest information you can get of brain functionality non-invasively. This is the first time our organization has such a big database and AI know how to look into it. We could see a big opportunity there. We continue, we plan, we already have the FDA breakthrough designation on delirium. We are working on the regulatory paths with FDA on delirium and continue investing in our stroke ongoing studies. I think that would wrap my presentation. Happy to take any questions. All right. Thank you.