Listen to James Piercy talk with Jeremy Holland, Chief Executive of Crainio, about their non-invasive intracranial pressure monitoring system.
Normally, when you need to monitor the pressure inside your skull, you have to drill a hole in the skull to put a bolt in the head to measure this. Crainio has developed a sensor which can be stuck to the outside of the head and monitor the pressure using infrared light and machine learning. The results are then continuously displayed in real time. This exciting new technology will reduce infection and could allow first responders to measure intracranial pressure at the scene of an accident. Jeremy talks in detail about the journey this technology has been on, and where they hope to be in the next few years.
| (0:09) James | Welcome to the latest podcast from the NIHR HealthTech Research Centre in Brain and Spinal Injury and today I'm talking to Jeremy Holland from Cranio. So Jeremy, Cranio (interesting name) obviously comes to do with heads. What is it? What is the tech? |
| (0:25) Jeremy | So Cranio has technology that enables the non-invasive measurement of intracranial pressure. So intracranial pressure is the key physiological vital sign of the severity of traumatic brain injury. So when you bang your head, your brain sort of gets hit inside your skull and it gets inflamed and it might swell up and it does that inside a fixed volume cavity and as it does that it gets bigger and it increases the pressure inside the skull. |
| (1:03) James | Okay and I'm guessing then the brain is getting a bit squashed under this pressure and that's going to cause more damage to the brain? |
| (1:10) Jeremy | That's right, so intracranial pressure in the first instance it's a sort of a measure of the severity of the brain injury but also it causes what's called secondary brain injury which is exactly that, it's pressure inside the skull damaging the delicate tissues of the brain and possibly forcing the brain against the hard edges of the skull and even pushing it down towards the spinal cord which can cause very serious damage to the brain. |
| (1:40) James | Okay, so obviously something really important to measure and disclosure, I had my intracranial pressure monitored after my head injury. I've had a little hole drilled in my head with a probe put in to measure so that's an invasive technology? |
| (1:55) Jeremy | That's right, so at the moment that is the only way currently used or commonly used of directly measuring intracranial pressure is for a neurosurgeon to drill what's called a burr hole into the skull and inserting a pressure sensor into the centre of the brain. It's a very simple physical process really just using a strain gauge to measure the sensor but it's extremely invasive, it's expensive, it's slow, it's complex and it has clinical risks associated with it. |
| (2:27) James | Yeah, obviously any hole you're going to make into somebody's head you run risks of infection and all sorts of problems so yours is non-invasive so how can we measure the pressure inside the head from outside the head? |
| (2:37) Jeremy | Right, so we use a magical technology called photoplasmography (of course it's not magic really). We use infrared light to shine a signal so we have a small probe that's attached to the forehead, temporarily attached to the forehead, that uses infrared light which shines at the scalp and at the wavelength that we use the scalp and the skull are partially translucent and so the light bounces off the surface of the brain and we can pick up photons that have bounced back from the surface of the brain and we can see the pulse wave in the arteries at the surface of the brain because that interferes with the infrared light that's being transmitted. |
| (3:23) James | Let me see if I've got the right idea, okay so you're shining something a bit like a tv remote control, (Jeremy: Yup) it's coming through the bones of my skull and it's interacting with the blood vessels inside but they're changing because the blood is *du-dum, du-dum, du-dum* pulsing through them. (Jeremy: Yup) Okay so the signal that's coming back is going to change depending on how much those vessels are being compressed. |
| (3:46) Jeremy | That's right, so as you can imagine, so we're looking at the shape of the waveform, of the blood pulsing through the arteries in the brain. As you can imagine, as the pressure inside the skull increases it squeezes the arteries, that means the blood can't flow as smoothly, and that changes the shape of the pressure wave. We pick up that pressure wave, we pick up the change in the shape (which is called the morphology of the wave) and we use machine learning to look at characteristics of that morphology and correlate those with intracranial pressure and so that's enabling us to say use machine learning to extract the intracranial pressure from the signal that we're getting from the arteries at the surface of the brain. |
| (4:30) James | So I'm guessing that between individuals that intracranial pressure is kind of, there's a normal range and you're looking for things that are kind of coming outside of that? |
| (4:41) Jeremy | That's right, so intracranial pressure is measured in the same units as normal blood pressure which is millimetres of mercury. Whereas your blood pressure you would expect to be sort of 120 or so millimetres of mercury, with your brain you would expect it to be between 5 and 15 millimetres of mercury would be the pressure inside your brain.
As it gets higher, doctors start to be worried and where the red lights really start flashing is when it gets above 20-22 millimetres of mercury and that's when clinicians would feel the need to intervene to try and reduce that pressure. |
| (5:19) James | Yes, some kind of intervention. Mine, I know from my notes got up to 27 and I was just given a pharmaceutical intervention so I was given a drug to modify the pressure and I can see on my graph that it came back down again. Other people might need some sort of surgical intervention to reduce the pressure. |
| (5:35) Jeremy | That's right, so there are various non-surgical interventions that you can use. I mean positional techniques, sort of moving someone so that their head is further away from the heart effectively. There are things you can do, hyperventilation, encourage very rapid breathing, you can encourage cooling of the brain, there are all sorts of techniques that you can use. There are pharmacological interventions which act in a effectively an osmotic way of extracting fluid from the brain. But then in extremists you can try and reduce the pressure by removing some of the cerebral spinal fluid which is a fluid in the skull or the most extreme intervention is something called a craniectomy where you're removing a bit of the skull just to give the brain a chance to expand without being within the fixed confines of the skull. |
| (6:29) James | Yes, give that space to grow. So you mentioned machine learning, so what you're measuring isn't actually the pressure but it's some sort of corollary of that and I guess the machine learning bit is understanding how the things that you're directly measuring relate to that intracranial pressure, is that right? |
| (6:45) Jeremy | That's right, so in order to, we've had to train our machine learning algorithm and in order to do that we needed to collect clinical data. So we've run a study in the Royal London Hospital on 40 patients, so these are patients that were in the hospital who had an invasive, the invasive probe, it's called a bolt, an invasive bolt applied. So it was measuring their intracranial pressure directly and we applied our probe at the same time and the signals, we were able to extract signals from our probe and then train the machine learning model so that it could correlate the features of the shape of the pulse with the change in the intracranial pressure. |
| (7:29) James | Yeah and I guess that's a fairly small study, that sort of number of patients, are you planning to do more to really kind of validate this technology? |
| (7:37) Jeremy | We are, we have two more studies coming up, running very shortly,one that we're doing at the moment. So we've improved the technology, we've improved the probe, we've improved the way that the light is emitted and the way it's detected and the sensitivity to noise and light pollution, so because we've changed the probe we're running a new study to retrain, to get more data to retrain the model and that's a slightly larger number of patients, so 54 patients and then, subsequent to that we will be running a further study to then demonstrate the accuracy of the device. So you need to train, so we're training the machine learning model but then we need to go through a process of, once it's been completed, once we have our algorithm that's been finalized, we need to show to the medical regulatory authorities that we're measuring intracranial pressure accurately, so that's a further study that we'll be doing in hopefully 2026. |
| (8:43) James | Sure yeah and just keeping it against that gold standard of the thing drilled in somebody's head. |
| (8:48) Jeremy | That's right at the moment, yes. |
| (8:51) James | Yeah great, so what does this thing look like? I think as well as you know that reducing that risk of infection and stuff, going to visit a patient with something sticking out their head is quite disturbing.
I know from my own family's remembrances that's not very pleasant thing to see, so is this thing just looks a bit nicer on the patient as well? |
| (9:08) Jeremy | It is, it's small, it's about the size of a couple of 50 pence pieces but it's flat, it's flexible and it's just applied to the surface of the forehead. There are many probes that are attached to the body like EEG probes, when you're having your heart rate measured, flexible, small, sticky, it looks and feels like one of those. It's applied to the forehead and can be there for as long as the intracranial pressure is measured.
It may be that you just want to measure it as a spot measurement but we can envisage people wanting their intracranial pressure to be measured over an extended period of time and we can show trends over that period of time. The probe is built to be put on the forehead for up to 48 hours. |
| (9:58) James | I think my bolt was in for about four days and I guess there's that critical period early on where we really need to kind of keep an eye on that pressure and see how things change over time.
So I guess the next thing to ask you is about money. I'm guessing that your device is a lot cheaper than these bolts? |
| (10:19) Jeremy | Glad you asked me that James. Yes we think so. So the NHS tariff for the invasive probe or the invasive bolt is £9,050 per patient. Now that's not just the cost of the probe but it's also the cost of the neurosurgeon and the operating theatre and everything you need in order to insert the bolt. We're placing our product at around about £100 per patient so that's getting on for two orders of magnitude cheaper.
And that means that we're looking at enabling the measurement of intracranial pressure in places that people wouldn't have envisaged measuring it previously. Making the measurement of intracranial pressure more commonplace because it's easier and cheaper and less risky and quicker to measure. So that gives us all sorts of opportunities to measure intracranial pressure. |
| (11:18) James | I guess all sorts of low middle income countries will really value this kind of technology just because it's much easier to use. |
| (11:25) Jeremy | I think that's right certainly and that's one of the markets that we're looking at but I think also it's in the UK looking at situations where you would not have expected to measure intracranial pressure. So what we want to do is get away from the operating theatre or the intensive care unit closer to the scene of the accident. So you can imagine ambulances, air ambulances, measuring intracranial pressure on the roadside and deciding what course of clinical action you want to take with this patient as a result of the measurement. |
| (12:01) James | That might be changing which hospital you take the patient to because some don't have such good facilities as others. If you know you're going to need neurosurgery you're very likely to you're going to go to a major trauma centre rather than a local hospital. |
| (12:14) Jeremy | Absolutely and there are many decisions that clinicians or attending paramedics will want to make. As you say where you want to take the patient but do you know the urgency with which you want to take the patient there. Notifying accident emergency to expect somebody with the following level of intracranial pressure and deploying their resources accordingly.
And as much of that as knowing that the intracranial pressure is not raised and so you don't have to rush them to the specialist neurotrauma centre. Maybe you can focus on some of the other injuries that the patient might have. Breakages or what have you, rather than worrying about what's going on in the brain which of course people want to protect because it's the most important organ that we have. |
| (13:04) James | Yeah absolutely and the speed thing I think is kind of crucial. I had my old drill in my head at half past one in the morning. That was about 12 hours after my accident because obviously I have to get transported and stabilised and scanned (Jeremy: Yeah). But, actually if you just snap a thing on the way to hospital you know how to prioritise things and you get much quicker. |
| (13:26) Jeremy | Absolutely and they have what they call the golden hours which are the first two or three hours after an accident. And interventions that you make in that time can have a much greater impact clinically than interventions that you make sort of further down the line. So the hope is that by measuring it early you'll enable smaller interventions and less invasive interventions to bring the intracranial pressure down and to reduce that chance of secondary brain injury that we were talking about. |
| (13:58) James | So Jeremy I guess you must have some helpful support and partners working with you on the project. |
| (14:04) Jeremy | Yeah well we work with many partners I mean the Royal London Hospital where we're conducting our study but I think our key partner is definitely, what's now known as, City St George's University. We're based in the Biomedical Engineering Research Centre there which is run by Professor Panicos Kyriakou and Cranios technology really is a spin out of technology developed at the Biomedical Engineering Research Centre and it was initiated by Professor Kyriakou and Professor Chris Uff the head of neurotrauma at the Royal London Hospital in 2016 and they felt that there had to be a better way of measuring intracranial pressure than drilling a hole in someone's head. And really Cranio is a result of that research activity that was set up as a result of that initial collaboration. |
| (14:49) James | So the Health Tech Research Centre I guess our role is to talk to amazing clever people like you and to kind of match to unmet needs. How has your kind of relationship with the centre helped the development of your technology? |
| (15:02) Jeremy | Well I mean it's been fantastic you've helped, when we talk about money, we're a start-up you know we ourselves need funding to grow and the HRC has guided us in our approach to grant bodies. We've raised significant grant funding to enable us to grow the technology and to further develop the technology and HRC has helped enormously with that.
They've also helped with access to additional expertise so there's you know people at the Brain HRC have got huge amounts of insight into brain injury and being able to access that and tap into that has been fantastic. One of the areas that you know as we get closer to the market we need to be more precise and more specific about the precise clinical pathways that we're going to be operating in, exactly how Cranio is going to be used. Who is going to apply it? Where, you know? What grade of nurse? What type of paramedic, you know? It's all of these detailed questions that Brain HRC is helping us to answer. Wnabling us to access the NHS more smoothly and more directly than we would otherwise do. |
| (16:20) James | Well we're very pleased to help in any way that we can on that journey. Where are you at now then in terms of getting this device validated and then implemented into use in the NHS? Are we looking another five years or it's going to be? |
| (16:32) Jeremy | No it's not as long as that I mean people say that it takes about eight years to take a medical device to market. Luckily we started eight years ago so we feel we're in the final straight. We have a as I said earlier a further what we're calling a feasibility study which is collecting data to train the machine learning model to get it a little bit more accurate than it is at the moment and then there's a subsequent study demonstrating our level of accuracy. But we're hoping to get onto the market in 2027 which, you know, time gallops along, it won't be long now that's certainly certainly the hope. |
| (17:16) James | Well we wish you every luck with getting this implemented it sounds like a fascinating new technology. So I've been talking to Jeremy Holland from Cranio do look up their website and find out more about this fascinating technology. Thanks ever so much Jeremy. |
| (17:28) Jeremy | Thank you very much James. |
