Hello friends,
In this episode of Tech Tuesday on The Incubator Podcast, hosted by Daphna, Mark O'Sullivan, PhD, CEO, and co-founder of NeuroBell, discusses their cutting-edge medical device aimed at improving neonatal seizure detection. NeuroBell, a spinoff from the Infant Research Center at University College Cork, has developed a pocket-sized, wireless EEG monitor equipped with AI-based seizure detection for newborns. This device is designed to be user-friendly for bedside staff and eliminates the need for extensive EEG training, making it accessible in hospitals with limited EEG resources.
Mark shares the story behind the development of the device, highlighting the challenges of acquiring EEG data due to its small signal and the importance of minimizing seizure burden in newborns. He explains how the device addresses gaps in care by providing real-time seizure detection in settings where traditional EEG monitoring might not be readily available.
Currently in clinical trials in Ireland and the U.S., NeuroBell aims to provide a streamlined solution that could be particularly useful in community hospitals and smaller settings. The episode wraps up with Mark's vision for the future of NeuroBell, which includes FDA submission and market release in 2025.
---
Find out more about NeuroBell here: https://www.neurobell.com/
----
The transcript of today's episode can be found below 👇
Daphna Barbeau (00:02.21)
Welcome back to Tech Tuesday. We're so excited this morning to have on Mark O'Sullivan, PhD. He's the CEO and co -founder of NeuroBell. This is a medical device startup based in Ireland, and they are developing novel technologies for seizure detection. So NeuroBell is a spinoff of the Infant Research Center at University College Cork, which focuses on translational research for pregnancy.
birth and early childhood. And we are talking about NeuroBell's flagship product today. It's a user -friendly pocket -sized and wireless EEG monitor with AI -based neonatal seizure detection. And it is currently in clinical trials, both in Ireland and in the US. Mark, thanks for joining us today.
Mark O'Sullivan (00:46.382)
Great, thanks for the introduction and for the invite.
Daphna Barbeau (00:50.25)
So tell us a little bit about kind of the inception story. Why was there a need for NeuroBell and the product you're going to tell us about
Mark O'Sullivan (01:03.512)
Yeah, well, I suppose my own background is I'm originally an electronics engineer, so didn't come into this with a clinical background, but always kind of had an interest in signals in general originally in the kind of the audio space. And then that's expanded to biomedical. And when I got hooked on the area of medical and just the of the drive and the satisfaction that comes up making.
technologies that can potentially improve patient outcomes. That was kind of my driver. So I decided to pursue a PhD with the Infant Center based in Cork in Ireland, which focused on the area of neonatal seizure detection and brain research. And during that opportunity, I suppose, during the PhD research, there was a big focus on developing technologies that were clinically useful. And every couple of weeks we got to meet with neonatal nurses, neonatologists, neurologists.
And it just kind of, suppose originally we went in looking at purely algorithms for EEG and for seizure detection. And then we kind of realized there's a bigger problem there of just access to equipment, access to expertise for placing the leads and for recording EEG. And I think it just kind of began to grow until a point where we realized, okay, PhD research is one thing, but actually making something useful at the end of this. There's no point in solving one very small
part of the problem. We wanted to develop a product or a system that would solve the bigger problem of can we have a bedside tool for EEG monitoring that can be usable by the bedside staff without having to do hours or days or weeks of training. And that also provides objective decision support. So that was kind of the thread that we kept on pulling on and it just took shape from there.
Myself and my co -founders Alison and Colm kind of help bring the technology to where it is today. And yeah, I suppose just going back to the clinical side.
Mark O'Sullivan (02:54.65)
From an engineer's perspective, it's a very interesting challenge because EEG is such a small, small signal. It's about 10 to 20 microvolts, which compared to your ECG signal, it's hundreds of times smaller in amplitude. So it's a really challenging signal to acquire. And guess that's why it's typically been such a difficult monitoring problem, because it does require a lot of due diligence in terms of placing electrodes properly and having
complex system that has all these different configurations but we believed that this could be achieved with kind of modern electronics to bring this down into a pocket -sized device. So that's kind of the inception of the idea. I guess it's born from my PhD and my co -founder Alison's PhD research. We co -founded the IP with the University and then split out with the University so we've always had a really good
kind of hand in the hospital system there, just being part of the Infant Centre, we had access to clinicians who could kind of constantly guide us in the right direction. it's been a big team effort.
Daphna Barbeau (03:55.598)
I love that. So I was so excited to read about what you all are doing. So I trained at a major academic institution here in the U .S. We had all the services, we had 24 hour techs monitoring EEG. And then I graduated and I went out into clinical practice and I was shocked to learn that actually most places didn't have anybody monitoring EEG at night. And there were some NICUs that didn't have the EEG capability.
you know, in real time at all. And this really severely affected which patients they could monitor and patients that had to be transferred elsewhere. So I can certainly see the need for product that I think we might have a lot of listeners who say like, why would we even need this? We've got EEG techs, but that's not the case actually in most places in the world and certainly in the US. What types of patients
Did you guys intend this for?
Mark O'Sullivan (04:58.146)
Yeah, so I suppose the device itself is a pocket sized wireless EEG monitor. It's a two button system, so in terms of the users, it's kind of any bedside staff.
The algorithm that's embedded on the device for seizure detection is trained on data for full term neonates. So that's kind of our initial indication for use is it's a H channel EEG monitor with integrated seizure detection for full term neonates. But the hope would be to expand that into kind of preterms as well as bigger kids. And that's kind of our plan so far.
In terms of kind of where we see it being used, I think you're dead right. The academic hospitals tend to have 24 -7 EEG techs. And I guess our hope with this device is that it can be used in probably some of the hospitals that don't perhaps have 24 -7 EEG techs or monitoring. We've had a lot of hospitals tell us that they're pretty well covered from kind of 9 to 5, Monday to Friday.
outside of that can be very difficult. Even in some of those hospitals, we've been told that there is delays getting kind of the EEG hooked up. And I guess this is where some of the new clinical data that's come out in the space has been quite informative. I think there's been studies from Toronto SickKids as well as other hospitals showing that the total seizure burden experienced by an ENA is directly correlated with outcome. So there should be a big focus on trying to ensure that we minimize seizure burden for newborns.
and if there's a one hour or two hour delay in terms of getting the EEG hooked up and getting the first set of eyes on it, that might unfortunately cause a big difference if that baby is seizing and obviously the challenging thing with newborns specifically is that their seizures typically don't have a clinical correlation so I think the data shows that 66 %
Mark O'Sullivan (06:42.868)
and neonatal seizures don't have any physical correlation so you are entirely reliant on EEG monitoring to accurately detect them. So it's a really challenging problem to try and solve and I guess hopefully with the bedside tool we're in to solve that.
Daphna Barbeau (06:59.19)
Excellent. it sounds like right now the primary use is kind of rule out seizures and babies where seizures are suspected. Or it sounds like this HIE population, perinatal asphyxia, where we know they're at risk and then monitoring for seizures in those first few days.
Mark O'Sullivan (07:16.992)
Exactly, yeah, I think in terms of indication procedures obviously HIU will be probably the main patient cohort but there's also stroke, infection, you know, also I think there's a growing kind
number of hospitals are looking at EEG monitoring postoperatively for neonates with congenital heart defects. So I know we're working with Texas Children's in that space in terms of looking at monitoring seizures postoperatively in newborns who require cardiac surgery. there's a growing list of indications where EEG should be used in neonates. That's driven by the recent ACNS guidelines as well that are kind of recommending
you should be using the units to provide the best quality of care.
Daphna Barbeau (08:04.046)
Tell us a little bit more about, you told us a little bit about the tech of the device, but logistically tell us, you what does it look like? Who's putting it on? How long does it take to put on? And how much training is involved for the people who will be doing
Mark O'Sullivan (08:20.248)
Yeah, no, so it's the technology. guess we focus a lot of our time in everything that you typically see on a computer screen with a conventional monitor in terms of all the settings of mapping channels to bipolar channels and setting up all of your impedance checks to make sure the signal quality is good. We do all of that natively in this small little pocket size device. So it's about the same size as a cell phone with a very simple two button interface. Press the on button, press the record button.
Daphna Barbeau (08:47.168)
It's literally red green, right? Easy to use, yeah.
Mark O'Sullivan (08:49.048)
Red, green, exactly. Yeah. And in terms of them, the device itself is the device that provides the seizure indication as well. So as the device records, it's constantly checking the data that it's recording. And if the AI algorithm...
detects that there's a high likelihood of seizure in the past 16 seconds or eight seconds, the device itself starts to flashlights to say, you should look at the data. So that's kind of the setup is ideally it should be grab this off the shelf, plug in the 10 electrode lead. So it's an eight channel EEG device plus ground and plus a reference electrode. So 10 electrodes in total. You can use it with whatever.
Electro leads you want to so if you want to use individual leads you can if you want to use a head cap There's a number of head caps that are compatible with the device as well in terms of who uses it. I guess that's kind
Daphna Barbeau (09:42.904)
Sorry, head caps you mean for like cool, cool, head cooling.
Mark O'Sullivan (09:47.222)
Even just for placing the electrodes, there's some nice new technologies in terms of caps that kind of have pre -placed locations for, you know, F3, F4, C3, C4 and the different sites that you need to place the electrodes at. So it's kind of up to the hospital in terms of what their preference is and then we'll be working with them in terms of.
Daphna Barbeau (09:49.016)
huh, okay.
Mark O'Sullivan (10:05.294)
making sure that that works with the device. But the device can be used with pretty much any wet electrode. And in terms of who uses it, I guess it's really going to be on a hospital by hospital basis. We have done usability studies with nursing staff. We've done it with neonatologists and with EEG technicians.
We did some recent research with Parkview Hospital Group in Indiana where we actually used it in transport. So we also think that this could be usable in transport. And we did kind of just a feasibility study of one adult volunteer and an EEG technician was able to place the 10 electrodes on the patient's head in, I think it was about nine minutes. So to go from an EEG technician sitting into the back of an ambulance that was in transport, so in transit down a
good quality EEG recordings coming out of all the channels. So that's kind of the scope. Obviously there's a lot of work to do just to try and get a training program together so that you know whether it's nurses or neonatologists or junior doctors that everyone feels comfortable using it.
Daphna Barbeau (11:11.386)
And do you envision this will fully replace someone reading the EEG entirely? How will we collaborate with our kind of our neuro colleagues, our neurologists, at least here in the States, for these types of patients?
Mark O'Sullivan (11:29.176)
Yes, no, it definitely won't replace the neurologist in the loop. guess the benefit here is hopefully that it will flag the...
pieces of data to look back on with kind of a bit more scrutiny in terms of areas where we've flagged that there is a likelihood of seizure, but also would hopefully improve workflows that when the device starts to flash to say that there is likely a seizure, that will be kind of a workflow improvement so that the neurology team can then be notified of this section of data that they should probably look at that now. And if there is a baby seizing or starting to seize that you can actually make that decision quicker.
That's kind of the workflow that we expect. The other piece there is, I suppose it's recording eight channels of EUG, so we tried to comply as well with the ACNS guidelines for multi -channel neonatal EUG. So it does make up one of the montages that's recommended there, which also means from the neurologist's perspective that
It with the normal CPT codes. So I suppose it's trying to perhaps bring a device that's somewhere between the amplitude integrated EEG, which is a two channel device, and a full EEG. So we don't see it replacing video, multi -channel EEG. And we don't see it replacing AEG, but there's certainly a gap in the middle there where.
perhaps AEG isn't enough. mean there's some data showing that AEG has a sensitivity of about 50 percent so I think there is room for improvement there and if an 8 channel easy to use EEG machine can provide a lot higher sensitivity and be kind of interpretable by
Mark O'Sullivan (13:04.212)
team and that hopefully improves collaboration between the neonatologist and neurology team as well and that should be a really good outcome and then for those patients who do perhaps have more complex issues they can move on to video EEG monitoring with know if you want to do full montage.
Daphna Barbeau (13:21.438)
And tell me a little bit about where are you in the process of the trials for the product? What do you have ongoing?
Mark O'Sullivan (13:28.952)
Yeah, so we've three studies that are either just about to kick off or it's ongoing. One in Cork, one in Indiana and one starting in Texas.
And the purpose of those studies is really just to demonstrate that this device does record clinically -grade EEG quality and to kind get us over the mark with that in terms of demonstrating that 8 tons of EEG from the NeuroBell device is the same as recording 8 tons from an ASUS device, albeit with much improved kind of setup and usability really. With retrospect, we'd be looking at that data for seizures and demonstrating the effectiveness of the algorithm in terms of
positive percent agreement, false detections, and just making sure that we hit the right mark on all those fronts. But the main benefit there really is just getting the feedback from the users, seeing what it's like in practice in the field. We're learning a lot in terms of, know, especially on the software side.
terms of the data so like the device itself is fully standalone it records EEG, it processes the EEG on the device itself and provides the seizure indications but all the data is streamed over Wi -Fi to a web application and that's where anyone can log in and view the data and understanding the kind of workflows there so I think that's really the benefit of these couple studies is showing just kind of the workflows of how the device is used and I think it would be kind of longer
term then looking at having a massive impact on outcomes and how we can structure trials around that to demonstrate improved outcomes for neonates by using bedside EEG and hopefully reducing seizure burden and as a result improving outcomes. So that's kind of the next focus for us.
Daphna Barbeau (15:12.15)
I'm thinking of a question that I hadn't thought about previously about the logistics, but you mentioned with these kind of storage and readouts and data. Is it able to be viewed then remotely the information?
Mark O'Sullivan (15:26.786)
Exactly, yeah. So our software team here have worked a lot on basically a web viewer that you can open in any web browser. So whether that's Chrome or Safari. And it means that once you have a login from the hospital and you're approved with the kind of the role based authentication, whether you're looking at that remotely or in the hospital, you can log in and see all your patients basically. So hopefully that should streamline the process as well, particularly for the neurology team.
And yet working a lot then with the teams on the cybersecurity side, making sure that the entire chain of data is secure from recording all the way through to where the data is stored and viewed. So, yeah, a lot of work with the IS team, the IT teams. But it's great. mean, it's all with the overall goal of having, you know, secure data. you know, there's a lot of sensitive information there. So it's making sure that all that's secure and.
Daphna Barbeau (16:08.142)
A lot of work.
Mark O'Sullivan (16:22.222)
that we also provide benefit to the clinical teams both from the NICU all the way to the Neuromuscular Department. And in particular, then if it is a smaller hospital that might be outsourcing some of that readout that they can have consultants or whoever might be reviewing that remotely under the same agreements.
Daphna Barbeau (16:41.07)
And what sort of feedback have you been getting either from the healthcare professionals and you must be interfacing with some families at this point. What are they saying?
Mark O'Sullivan (16:52.77)
Yeah, it's been really positive, I think, from the bedside users. think in particular, nurses are excited about the possibility of having a device that works at the bedside that provides objective decision support. We've had a couple anecdotal comments about they've had suspicions about a baby that might be seizing, but there's a big call to be made whether they make a call to neurology to get an EEG down. In a lot of cases, that's not an easy decision to make, in particular when the EEG department isn't part of the same hospital and
service level agreements between different hospitals and just making the simple decision of should we get this baby on EEG isn't a simple decision to make. So I think the nursing staff that we've spoken with across both Europe as well as the US have been excited about the possibility of that, that there's no longer this need to make a big decision about whether we're getting EEG or not. There'll be an EEG device in the closet that you can pull out and put on the baby, record for 30 minutes if you want
and then make a decision based off that. I think the nursing team are really excited and similarly the neonatologists. I think having that kind of objective decision support without being 100 % reliant on potentially an external department is exciting. For the neurology team, mean, think across the US there's a shortage of EEG technicians, which has put a strain on EEG services, in particular one of the hospitals
They have a team of kind of five or six EG technicians that are spread across neonates, pediatrics, adults, and they're also kind of rotating EG machines like it's musical chairs. So it's a good opportunity here for them to be able to kind of streamline their processes. If the NeuroBell EG monitors stay at the NICU, it means that they are not pulling their staff across, you know, different sections of the hospital and wheeling EG monitors across different parts of the hospital.
but they still get the benefit of being able to review it remotely from their offices and I suppose make sure that any newborn who does require EEG or is having seizures is getting the right monitoring and treatment. So I think the feedback so far has been great. I suppose we're trying to satisfy a lot of different potential user groups from nurses to neonatologists to EEG technicians and neurologists. So it's a difficult piece there, but it's very collaborative and so far I think we're on track to hopefully improve the kind of across all.
Daphna Barbeau (19:11.82)
I love that. you know, I think what you're describing is going to become actually more prevalent. know, hospitals are not moving to include more services, right? They're trying to include less services. So I think that, you know, will more and more hospitals will be finding themselves in this exact dilemma that you describe, even those that have the full range of services right now.
Mark O'Sullivan (19:36.216)
Yeah, I think it's there's an interesting kind of divide there where the guidelines and the clinical guidelines are recommending more and more usage of EEG in neonates, but hospitals are actually trying to streamline services as well. So it's kind of putting further challenges, I would say. So unfortunately, both of those pathways are diverging as opposed to converging.
Daphna Barbeau (19:42.527)
That's
Daphna Barbeau (19:55.135)
And have you been able to meet with any parents?
Mark O'Sullivan (19:58.222)
We have, we've met with a couple of parents in Indiana, particular, kind of one or two parents who've had newborns who've gone on to have seizures and kind of just talking to them about the experience of any other conventional EG monitoring put on. Obviously there's kind of, it's a big unwieldy system that gets brought to the bedside. The parents don't know what it is.
Daphna Barbeau (20:21.632)
It's very intimidating, I
Mark O'Sullivan (20:22.878)
it's very intimidating and it prevents any kind of you know the parents from touching or kind of holding the baby. I think that is one piece that we'd like to look at in time is potentially with how portable the system is and can the mother still be able to hold the baby while the EG monitoring is ongoing. But I think like even going through the consent process you know it's always such a difficult piece and you know I'm a bit detached from it being an engineer by background but
and the amount of respect I have for, I suppose, the research nurses and teams that have to go and obtain consent for these studies and talk to the parents who are obviously going through a pretty traumatic time and trying to explain, know, this is hopefully for the benefit of their child, but also children to come hopefully.
And yeah, think in general, parents that have spoken have just been appreciative of, OK, there's a new device coming that might be able to make this whole process of the EEG monitoring less intimidating, but also more streamlined. think parents that we've spoken to have kind of mentioned that you can be sitting at the bedside and waiting for quite a while.
and they know that something is due to be happening, but they can't quite put their finger on it. I think it's, hopefully this is a nice, similar tool that if there's a mention by the nursing team or the bedside team that, okay, we're going to start some brain monitoring to see if there is any abnormal brain function, that that's not now waiting an hour for someone to come from a different apartment or for calls to happen, and instead it can happen immediately, and hopefully provide some reassurance.
Daphna Barbeau (21:32.462)
Mm -hmm.
Daphna Barbeau (21:56.184)
So what does the future of NeuroBell look like? What's the long -term vision?
Mark O'Sullivan (22:00.954)
Yeah, so I suppose we've been developing the technology for the past eight years in essence, guess myself and Alison both started our PhDs in around 2015, 2016. Then we went on to do research in the space, but always tried to push this technology out. And we're finally out of stage now where it's in clinical trials and the aim would be to be filing with the FDA towards the end of this year, which means hopefully middle of next year to have the device on the market and to be.
getting it out there in use, not as part of studies kind of with conventional EEG on -site, but instead studies where it's just the NeuroBell device as a standalone device improving outcomes. So just really eager to engage with US hospitals that potentially would see the benefit in having kind of a streamlined EEG process, a system for recording EEG quickly and providing automated.
seizure detection and I think that spans kind of your big level four tertiary hospitals as well as level three and level twos. I think we're just eager to make sure that this is a system that's going to work for all different hospital types as opposed to having a system that is a bit out of touch with you know some of the smaller hospitals so trying to find processes that work for everyone and so just really eager to get more feedback more trials I've been running and hopefully on the market in the middle of next year.
Daphna Barbeau (23:19.726)
And so if people want to learn more, they can go to your website, www .NeuroBell .com. And it sounds like you are hoping to get more hospitals on board, more people trying the product. So what's the best way for people to contact you if they think that this is very intriguing and they'd like to give it a
Mark O'Sullivan (23:41.294)
Yeah, well, suppose either through the website or else my email is America .newrville .com as well as that we should be kind of on the door of a lot of the conferences this year. So the Delphi conference coming up in September, you'll see us there. We were at PAS earlier this year and also the International Nubian Brain Conference at the start of this year.
So it's kind of an exciting time in terms of just being out and talking with people both email or through the website or else grab us at any one of the conferences. We're always eager to talk to anyone in neonatal care, regardless of what sector that is. It's always good to learn kind of the workflows and people's experiences.
Daphna Barbeau (24:19.928)
Well, Mark, thank you for all the work that you and your team are doing to fill a need. And thank you for coming on to tell us about your team and your company. It's been a pleasure chatting with you today.
Mark O'Sullivan (24:33.284)
Great, thanks a lot for your time. Appreciate it. Talk to you soon.