Hello everyone, I'm Javier Lorenzo. I am the co-chair of the in-training section in the Society of Critical Care Medicine. I am really thrilled to invite you and have you present for our crosstalk neuroscience with the topic of neuromonitoring, multimodal neuromonitoring outside of the ICU. I'd like to introduce our two speakers, Dr. Sahar Safar and Dr. Arti Sarwal. Dr. Safar is a neurointensivist and a clinical neurophysiologist and the associate medical director of the MGH Neuroscience ICU. She's also the director of the MGH Harvard Neurocritical Care Fellowship Program. Dr. Safar performs comparative effectiveness and outcomes research in neurocritical care and clinical neurophysiology with a specific focus on the comparative effectiveness of ICU EEG monitoring and antiseizure treatment strategies. Dr. Sarwal is a professor of neurology at the Wake Forest School of Medicine. She is a section chief for neurocritical care and the medical director of the Neurocritical Care Unit at Wake Forest Baptist Medical Center. Her research interests include point of care applications of ultrasound in the intensive care units, neuroultrasound for assessing cerebral physiology and cerebral hemodynamics, and ventilator-induced muscle dysfunction in neurocritical care population. Welcome both, welcome everyone. I do not know very much about multimodal monitoring, so I was really excited when I got the chance to moderate the session. And I'm gonna start with the first question that comes to mind. What is multimodal neuromonitoring and how is it currently used? Dr. Sarwal. So thank you, Javier. It's great to meet you through this forum and honored to be a part of this panel and discussing something that I'm very passionate about. So I think when we traditionally talk about multimodality monitoring in the ICU, we probably are doing it in many ways, whether it's a neuro ICU or a other subspecialty ICU or a med surg ICU. So when we talk about different ways of looking at the exam, integrating that with imaging, integrating that with systemic hemodynamics, whether it's invasive or non-invasive, integrating that with perfusion assessments and basically integrating multiple modalities into our clinical decision support algorithms, that is what I take as multimodality monitoring. Whether you are taking care of invasive monitor, say like a ICP monitor, or you're looking at just a pulse ox waveform using a peripheral pulse ox meter, each of these components for me is a multimodality. Sahar, what do you think? Absolutely. I mean, I think even at the basic level, when you walk into an ICU and you look at the monitor where you have the patient's heart rate, pulse ox, respiratory rate, that in itself is multimodal monitoring. And then bringing the neurology context into it, we add in intracranial pressure monitoring, and then there's a range from invasive to non-invasive. So non-invasive would include our EEG monitoring. We have non-invasive measures of measuring blood flow, such as NIRS or infrared spectroscopy, a transcranial doctor, which I know Dr. Sarwal can shed light on a lot. And then it ranges all the way to the invasive side, which includes intracranial pressure monitoring, brain tissue oxygenation, microdialysis. And so integrating all of these neural multimodal monitoring with systemic, so not on its own, but with your systemic monitoring of your heart rate, of your blood pressure, is sort of the comprehensive multimodal monitoring approach. And I'll give you a little extra perspective on that, how I teach it during my sessions. So I think what I regard is, think of your house as a structure. In the house, there are basic elements. One is the actual architectural structure, that is the cement on the walls, which is our neuroimaging, if I just consider the brain as an example. And then there is plumbing, which is the cerebral blood flow. So there are ways to look at plumbing. Then there's electrical circuit that needs to be assessed, which is in our case, EEG. And then there is your chemical structure of the house, which is neurotransmitters. So when I teach multimodality monitoring, I'm basically teaching an integration of structural imaging, plumbing, cerebral blood flow assessment, electrical assessments, and then neurochemical assessments. And I look at the same way from a heart perspective or any other organ perspective. I really liked that analogy of a house, the different components that make a house. So there were a lot of neuromonitoring modalities just mentioned. And I wonder if you can maybe describe the top most commonly used and what they measure, how they measure, and how they're placed, whether it be an invasive monitor or non-invasive, how do they physically live in the surrounding of the patient? Maybe I'll ask that for Dr. Safar. Yeah, sure. So I think the most common forms of neuromonitoring are non-invasive form. And I think the most common of those being continuous EEG electroencephalography monitoring, which I think we're seeing more and more, not just in the neuro ICU setting, but outside of the neuro ICU setting. And the way EEG monitoring is done, again, you can get spot EEGs, which is you get a one-time EEG that looks at brainwaves for an hour, and then all the way to continuous EEG where you can monitor a patient continuously for several hours or even days. And this setup, it's a bedside setup done by EEG technologists. And often, depending on the ICU and the setup, you can have the EEG reading screen itself in the room versus a central place where the neurophysiologists review the EEG. So I think that's one of the most basic, or not basic, but the most common non-invasive form of EEG monitoring. Other non-invasive forms of monitoring include transcranial doctors and TCDs, which are used to measure cerebral blood flow, vasoreactivity, particularly in patients such as subarachnoid hemorrhage patients where we use TCDs for assessment of development of vasospasm. TCDs are more in a moment of time, so you don't get continuous data and are usually performed on, depending on the disease etiology. So if it's a subarachnoid hemorrhage patient, we usually perform these on a daily basis during that timeframe where they're the highest risk for cerebral ischemia. And then sort of the more, these are, I would say these count amongst the common non-invasive one. Another one to add here is actually pupillometry, which I think is gaining traction even outside of the neuro ICU and the medical ICU setting. And we don't always think of it in terms of a multimodal monitor, where you have numbers on screens, et cetera, being thrown at you, but we're learning that it can be a very helpful tool in identifying early pupillary changes, that can be a warning sign for impending cerebral edema or impending herniations, or sort of helping in picking those up early. And then sort of switching gears to the more invasive monitoring, I think the most common invasive monitoring is the intracranial pressure monitor. That in itself, there are a variety of options. You can have monitors directly into the ventricles, so external ventricular drains, do intraparenchymal monitors, subdural, epidural monitors. And I think the most commonly used are the EVDs, external ventricular drains, and the brain parenchymal monitors for intracranial pressure. Common indications are trauma, trauma being the most common, subarachnoid hemorrhage patients are another population. And then I think less common, but where the interest is growing is using brain tissue oxygenation monitors, which are also invasive, go into the brain parenchyma and microdialysis, which measures brain metabolism, including lactate by ratio, and then finally depth electrodes to measure deep intracranial electrical activity. And I know I've said a ton of monitors, maybe I'll let Dr. Sarawal sort of weigh in as well around the invasive versus non-invasive. I think the one most important thing to highlight is that with this monitoring, perhaps you can cover this in just a little bit, is that what's most important is not following just one value or going after what one parameter says, but it's really the integration. So trying to combine as much of this data as possible to really provide goal-directed treatment for patients. So I think I'll kind of follow my house analogy here to add to the monitoring systems you've talked about. So then let's start from the basic. I think when we examine a patient, using a standardized score like a neuro check, a GCS score, a force score, that is a new monitoring tool as well. You can extend to a device by using papillometry like Sahar said. So that exam, when you're using pain scale assessments, whether it is a behavioral pain scale or a sedation scale or a delirium scale for CAM-ICU as an example, these are all neuro monitoring tools. So that would be the exam part. On the plumbing part, obviously you can do static as well as dynamic blood flow assessment. Sahar mentioned transcranial dopplers. We can do static assessments with CT perfusion. We can have xenon flows. Then on the electrical side, she mentioned EEG. BIS monitors are a form of kind of electrical as well. On the plumbing side, on the non-neuro ICUs, we use NIRS a lot. The NIRS, that's a very common use. Then on the neurochemical side, we can use micro dialysis in the brain. I think we are one of the few organs there where we can actually measure oxygen going into the blood or the brain parenchyma. We can measure the pressure of the brain parenchyma and we can measure the metabolism of lactate, thyroid conversions in the brain parenchyma as well. So I think in addition to your pulse ox, hypnography, these are kind of every modality ranging from exam to invasive is kind of a, it's a whole zoo of a Disneyland of monitoring devices that we have at our disposal. So I'm familiar with these neuromodalities that give you like a one-time variable, like the ICP when you transduce an EVD or a velocity in a TCD. But these macro dialysis, I mean, is this a catheter that is actually giving me a continuous variable? Is this a catheter that it takes some time to get some feedback from that? How does that work? I'm so fascinated. So micro dialysis is actually not continuous. So what you basically do is you take, there's a micro catheter that is placed at a certain pre-agreed upon location. Typically it will be at the gray-white junction at a particular depth in the area of interest. So if you have a TBI patient, you want it in the area where the patient is at risk for DAI. And then you basically take that sample and you analyze that in an analyzer and that'll give you lactate-pyruvate ratios. So it almost gives you a assessment of patients aerobic versus anaerobic metabolism and adverse lactate-pyruvate ratios tell you that cell death has occurred. So it's a marker that that particular area has seen cell death. It's afterthought and you can't do it continuously because it depends on how often you can get samples. Typical places which do micro dialysis will take hourly samples or queue for hourly samples. It's still a research modality and that I think static nature has not really allowed us to use this as a clinical modality at this point. And also the fact that it's afterthought. You detect damage when it's already happened rather than at risk territory. So say TCDs are cerebral blood flow parameters like CTP that actually gives you an index of what's at risk. So it allows you to create that as a tool to prevent injury rather than just prognosticating that bad injury has happened. I think one thing, oh, sorry, go ahead. I was just gonna make the comment that I guess that is the point of all these monitoring is to prevent secondary injury, to prevent... We can't do anything about what already happened on the field or the initial injury, but we're here to detect and mitigate secondary injury. Go ahead. I did not mean to interrupt you. No, what I was gonna mention is I think more the brain tissue oxygenation, which is also not utilized as often can be a more continuous marker of regions of perfusion as well as metabolism, as that provides more real-time numbers and monitoring. So I think since we brought up the idea of static versus continuous, I do wanna highlight maybe just the continuous part because there is one, I would say, advantage we have in brain compared to many other organs, just like you can get continuous pulse or continuous hypnography. The things that lend themselves to continuous brain monitoring are EEG. That's the best non-invasive tool. We have invasive EEG as well. Then the blood flow monitoring is NERSC, is a continuous monitor. The second continuous monitor, TCD actually can be done on a continuous basis as well. So there are headbands available where you can have continuous three to four-hour monitoring that is being proven to be safe. Brain tissue oxygen is a continuous monitor. ICP, whether it's a brain parenchymal monitor. So EVD would have to be transduced, but we have fiber optic-based continuous brain parenchymal monitors that can give you continuous ICP waveforms. So oxygen, ICP monitors, blood flow monitors. We have Hemidex, which is a continuous brain tissue flow monitor. Am I missing any other continuous monitors? No, I think what I can add to this is that we can actually place many of these monitors using the same port. So it's the same housing device through which you can put in. So it's not like you have to drill multiple holes to get each of these devices in. One device can carry a brain tissue oxygen, ICP. Brain temperature is another thing that we can measure continuously, along with the depth EEG electrode. All of these can be housed sort of in the same device and casing and allow for continuous monitoring of all these various different chemical electrographic markers. Does it matter where in the brain you put this bolt or is there a particular area of interest where you wanna make sure you have this in the right area? That's a great question. And I think there's two schools of thought around this. One school of thought would say, let's target the area where you have the most pathology. And the other will say, well, we already know that place is affected. Why not look at an area where we want to make sure we're preserving as much as possible? And I think part of this is also limited just by logistics and safety of placing the device. So in general, the most common location where we put a lot of these monitors is in the right frontal lobe because that's where you have the least real estate. But of course, if majority of your injury is on the left side, I think in general, what we try to do is if there is injury on one side, which is not devastating, and you know that there's area to preserve, then target the area where the actual lesion is, unless of course there's challenges in approaching it. I don't know, Dr. Sarwal, which school of thought you belong to, but. I'm actually the non-invasive school of thought. One challenge I always have with invasive monitoring devices, having trained in them is two thought processes there. One is when you're using invasive devices, you're looking at the parking lot. Are you interested in the parking lot? Are you interested in the blood flow that is the traffic going into the parking lot? And I think my thought process is that you can't isolatedly pick up, if that's a word, a clinical algorithm that only looks at the parking lot or the traffic flow. I think it has to be both. And there I feel the invasive devices have kind of failed us because they're losing the bigger perspective of the big highway traffic that is going to clog this parking lot. So in that sense, I have become a little more holistic advocate for saying non-invasive TCDs on both sides actually give you that information. I obviously am biased. You will hear me beat my drum for TCDs pretty much every chance I get, but I do think there is some value in assessing the parking lot as well. And the second challenge that I mentioned there is the Heisenberg principle. And the challenge that invasive devices have is the micro nature of what you're studying. It really gets to that small level that by inserting the device in a damaged tissue, are you damaging the tissue as you're doing it? I mean, if you've used invasive devices, we all know a significant amount of these can cause very insertion hematoma. There is some degree of inflammation and edema that comes from that. And the dam, the circle of the tissue area that you're studying with that, does that really allow you enough reliable information of things not caused by the monitor itself to impact patient's management and secondary brain injury? So I think that's kind of my school of thought that look at the parking lot, but look at the traffic going into it as well and don't get so siloed into what that limited information is going. Like you said, don't look at a number, but understand the limitation that you're only getting a two centimeter square area of that monitor to inform. And should that really be informing my clinical judgment on a bigger perspective? I love these analogies. And I love that principle that just the fact that you're observing it at such micro detail, observation itself can actually cause changes in what you're observing. I'm curious, who gets these types of monitors? Do we put this on all patients in the neuro ICU or are we more, do we try to triangulate the patients who are most likely to benefit and who might those be? Dr. Sarwal, I'll start with you. So I think the biggest neuro monitor in any ICU is your patient's neuro exam. If I'm sitting up and complaining about hospital food, I do not need an invasive neuro monitor. I think that's plenty neuro monitoring for me. So the patients that we get, or my approach to neuro monitoring, whether it's invasive or non-invasive is, one, my traditional pillars of exam have failed because the patient doesn't have an exam as a result of the primary brain injury or the patient's exam is confounded by my clinical paradigms. You know, patient has required seizure treatment requiring significant sedation or has ARDS that is requiring paralysis where my ability to monitor something that would cause secondary brain injuries extremely limited. There are paradigms that have been suggested by guidelines like the Brain Trauma Foundation guidelines recommend a particular GCS target and a particular CT scan characteristic that would lend itself to patients who will actually benefit from invasive ICP monitoring. I think what you have to decide is extra information always is not the best approach, I think. So go for the neuro exam. If the neuro exam is limited, which patient can you benefit by justifying the risks of your monitoring? And when I say risk, it doesn't mean that only invasive monitoring, even non-invasive monitoring is a risk. A little knowledge can be a bad thing sometimes. So that's kind of my approach of who I monitor. Sahar. Yeah, I think it's fairly similar. I mean, I think at the most basic level, as you said, it starts with the exam. If you have an exam that's reliable, you're golden. And then the next step comes, if you have an exam but the exam is out of proportion with what you would expect. So if I have a small frontal stroke where I would just expect that, I would only expect to see some focal weakness, for instance, but the patient's exam is out of proportion, they're more encephalopathic than I would expect them to be, you know, their weakness is more profound than I would consider EEG monitoring, for instance, is there something else going on that we're missing? So if the exam is out of proportion, and then of course, the comatose population where you really it's hard to follow, and it's also hard to follow if you're doing an intervention, you know, to follow that accurately. And so so those are sort of the gradation, you know, exam based gradations, then the disease based, I think, for instance, out of disease populations, TBI is perhaps the most common where there are recommended brain trauma foundation guidelines, as Dr. Sargon mentioned, for considering ICP monitoring when their GCS is lower when they have certain imaging findings or certain hemodynamic findings. After that, I think it really just depends on you know, what the disease processes and what the clinical picture is, I think the one disease condition where we probably don't do invasive monitoring, or at least at our institute, we don't do invasive monitoring or is acute stroke patients, for instance, where you know, we would consider non invasive EEG or transcranial doctor monitoring, but that's not a population where we would usually consider invasive monitoring, acute ischemic stroke patients. I see. So I wonder, what is the evidence for these neuromonitoring? Say we have that one patient that we all agree, the exam is limited. They're not complaining about hospital food, instead, they're comatose. There might be a severe TBI. Room A, we put neuromonitoring on room B, we manage with, you know, conventional traditional approach. Is room A going to do better? Because now we have all this information from multimodal neuromonitoring, but what's the state of the literature on its effectiveness? Yeah, I think that's where the controversies arise, right? Because, you know, unfortunately, the state, you know, there are a few studies or large trials or, you know, you know, clinical trials that have looked at least at intracranial pressure monitoring, for instance, that's the one that, you know, we've done, you know, have the trials, you know, there's the BESTRIP, the chestnut trial. And while monitoring, you know, does show that you can get the ICP, lower the ICP outcomes that were, in fact, not found to be significantly different between those who are monitored and those who are not. We have an ongoing trial, the BOOST-3 trial, which is looking at brain tissue oxygenation monitoring in conjunction with ICP to see that, you know, how that impacts outcomes, but that is an ongoing trial. And then I think with EEG, there are no trials, there are larger observational studies that have shown that EEGs, EEG can, for instance, help predict delayed cerebral ischemia a little bit, you know, earlier than some of the other modalities, but again, how that impacts outcomes is yet to be determined. So I think for most of these monitoring devices, the measurable impact on outcomes is yet to be defined. Before I switch, I'll let Dr. Asarwal take over, my little, my one other thought around this is that I think the challenge really is, is the way these trials are approached. And, you know, when you look at the ICP trials, it looked just, you know, you're doing just like targeting that single number. And I think that's where we fall short, where we're not taking the entire picture, patient picture into context, and, you know, going after just that one number. You know, if I had, if I see a patient whose ICP is high, but this is, this is, you know, not a comp, you know, not an example you see, but, you know, if the ICP is high, but their exam is fine, I wouldn't treat them, right. So, so it's really integrating and doing trials and studies that integrate all the monitoring that we have, as opposed to comparing, you know, just one modality and one number versus everything else. Yeah. Javier, it's all voodoo. Well, clearly not. I mean, there's so much infrastructure and so much investment in some of these neuro monitoring. So something has to be working. And I know that at least we have really good safety data for, for a lot of these neuro monitoring. And it's good to know that at least we're studying its efficacy as, as we, as we chat, but there's, well, is anything that you want to add to, to that answer? Yeah, I think there are two things I would add to that. Sahar really kind of hit the point home. One thing I will add is, I think we're underestimating how new are we to understanding neuro monitoring and understanding how to integrate that into our decision support algorithms. That's number one part. If you really look at the history of neuro monitoring, it was 1960s when, you know, people like Lundberg basically started putting monitors and everybody who came into the ICU and basically observing it. That study literally cannot be reproduced to have, you know, if you really look at the first ICP monitor study, it was 58 patients or something, if I remember the kind of some ballpark of 50 patients and all comers, you know, we didn't have a CT scan at that time. So we don't know if the patient had tumors. It was all clinical judgment, whether the patient has ICH or patient eventually got decompressed and said, oh, that's a tumor or, oh, that's a hemorrhagic stroke. And that's when we understood that patients above an ICP of 40 kind of didn't do well, patient less than 20 didn't, did relatively well. And patients between 20 and 40 were kind of all over. This is the kind of data that we are working on and the challenges given the regulatory challenges of, you know, today's day and age, that study cannot be reproduced. So our field is limited by the fact that I cannot do a population-based study of what is a normal ICP number. That's one challenge that I want to highlight. The second part is of all organs, brain is one organ that is a self-sustaining organ and a self-protective organ. What does that mean? We are the only organ that has the most robust auto-regulation and an infrastructural skull that protects us. Why does that impact? We do a study of induced hypertension in subarachnoid hemorrhage. We do a RCT, the most well-designed RCT. I think if I remember the only one that I know robustly done was about 40 patients. The challenge with that is, and Sahar kind of alluded to that, we're only looking at one particular thing and trying to translate that into a brain that, number one, is enclosed by the skull, has four different components going on at the same time that it's trying to have a self-protective algorithm. The brain is trying to auto-regulate itself. The brain is trying to maintain its blood flow. So designing a trial in induced hypertension in a subarachnoid patient cannot be completely translated until you have studied auto-regulation. Are the patients who are auto-regulated doing the same as a patient not auto-regulated? Or patients who have neurocardiogenic complications, the subarachnoid hemorrhage causes decreased cardiac output, at the same time as using induced hypertension, did you take into account the cardiac index? Did you take into account their CPPs as manifest by their perfusion? Was the systemic perfusion being addressed by the cardiac output? Are these patients oxygenated at the level that the brain is happy? So I think brain is so complex in that matter that designing a trial that only looks at one particular thing is, although as physiologically based as possible, cannot really justify what we would call as true evidence-based medicine. The last thing I would say is that has been challenging in producing the class one evidence that we have produced, say, in a drug trial. We do have class one evidence, you know, aspirin prevents stroke. The number needed to treat a patient for good outcomes for thrombectomy is like two. We have an NNT2 kind of a trial in our specialty. For neuromonitoring per se, the last challenge that I would describe is, look at the amount of data that we're looking at. We're looking at data integration, we're looking at data display, we're looking at data analysis in a form that actually helps us make meaningful real-time changes by bedside to design such a trial. That hasn't quite translated to a pulse ox, you know? So I think those are the three challenges I would say have been the challenge behind producing what we call as evidence-based medicine behind neuromonitoring. And one thing I would add to that is also, you know, where it's different disease subgroups or categories, right? There's TBI, there's SAH, there's, you know, who do we go after and, you know, who should we start with? And once you start designing trials which are restricted to one population, then you run into trouble with numbers and recruitment. So those are other challenges in designing some of these studies. Acute brain injury is not one disease. I mean, subarachnoid induced vasospasm is very different than a convexity subarachnoid causing vasospasm in TBI. They're just not the same disease. Right. So as you guys both mentioned, the challenge and the controversy of data integration, I wonder current states when you both use these neuromonitoring modalities, is it up to the clinician at bedside to integrate what we're seeing, the ICP, the cerebral blood flow, the pulse oximeter, the tissue oximeter? We know from, you know, cognitive scientists tell us that decision-making degrades the more data, the more inputs we have because we can only pay attention to so many numbers. So who is integrating this? Is it really the human at bedside? Should it be the human at bedside? Or maybe there's some other algorithm or analytical tools that are used clinically or that are being developed. What is the current state and what should the state be? Maybe that's a better way of asking that question. Dr. Safar, what are your thoughts on that? Yeah. I mean, I think the current state for the most part is it's the bedside provider. And I think that's sort of, you know, and this is part of the challenge to even do population-based studies, as Dr. Sarwal mentioned, that it's a bedside provider. And because of, you know, the way the data is all over the place and, you know, the way that, you know, there's guidelines for something, but not for others, there's going to be practice variation. And so, you know, someone who's very into multimodal monitoring, like myself, if it's my patient, then I will look at all the numbers, integrate it, you know, and, you know, try and approach it in sort of a more wholesome approach. But I think it varies on people's, you know, comfort level, but also how much, you know, they believe it. Because again, like we've talked about the data sort of all over the place. So, I mean, in my mind, given the current status, there's widespread practice variation, it will vary from provider to provider, from institute to institute, some centers are doing, you know, micro dialysis, some are not, some are doing brain tissue oxygenation, some are not. So, we haven't even standardized it, you know, at sort of a national or a global level. And so, the next steps, I think, for me that are most important are trying to do these population-level studies, whether it's clinical trials, or whether it's larger comparative effectiveness studies, really utilizing the sort of natural variation that we have across centers, providers, you know, and maybe perhaps pragmatic trials are probably the best way to really answer this question. But I think there's a lot yet to be answered for us to be able to have a more standardized approach. I think EMR, to some extent, as much as we hate our EMRs, can be that solution. I think we've kind of seen the seedlings of that. So, if you look at, you know, your bedside monitors in the ICU, I think from a display perspective, I think we're doing a better job of multimodality monitoring compared to, say, you know, when I was in med school. In med school, we literally had to take out rhythms at that time, and if a patient had a swan-gone scatheter, we literally had to print out those waveforms and kind of, you know, analyze them almost every hour. That was the core of my ICU rotation in med school at that time. But at least we have the, you know, the benefit of having a bedside monitor that actually shows those waveforms. Now, the challenge, obviously, is training the human mind for pattern recognition. So, I think this is where technology needs to come to our rescue. There have been early movements towards that. At least on the neuro side, we have early multimodality monitors, at least two or three commercial available, that integrate all the data in a manner that is digestible to human mind, in a manner that is applicable to clinical algorithms. You know, if you give me all that data every two weeks, and every two weeks, I get to know what the patient's auto-regulation state was two weeks ago, it's not going to be helpful. So, I think it has to be compressed in a form that is not just usable for me, but also usable in real time. You know, once a day rounding on a patient that is actively having dynamic compliance issues, you know, from six hours to six hours is not enough. So, I think EMR is early, like, I'll give you a small example. Looking at CPP, you know, just measuring CPP every time you pass by the patient's bedside is not optimal. So, having EMR display you CPP numbers and create some smart alarms, because the part of display and using them is not just about displaying them and giving you something, how often should I be seeing it? You know, am I going to have an ICU doc now sitting by the bedside monitor looking at it 24-7, see, oh, when does the value go above this threshold? So, displaying it and then analyzing it in the manner, in fact, in the manner that human eye can recognize that, and then creating some sustainable infrastructure or framework where smart alarms allow me to say that I'm going to respond to it when it gets to this, you know, ICP is more than 25, more than 20 minutes, CPP drops to this level, use EMR to trigger my response. So, complicated answer. I think we could talk about ours on that. Yeah, I think it really is. Sorry, all I was going to say is, I think this really, this is sort of the future, you know, I think a place for AI application, really. Once we figured out, these are the numbers, this is how we target, this is how we integrate, then the next step would be, let's bring in AI so we can do this real time as opposed to sort of, you know, playing catch up. Yeah. I was going to do a cheesy quote there. I think multi-modality monitoring in the ICU, whether it's new or not, needs an iPhone movement and Steve Jobs. I think there is a machine learning potential here because there's so much data and I think that we can only keep so many things in our head. We're going to need to, and also the more data we're producing, the smarter our algorithms will get in predicting who, you know, who's going to decompensate. But I guess I also want to ask, you know, the cost of these monitors, and I hate to say this because, you know, we live in a healthcare system that is ever so more expensive, but I can only imagine how expensive, you know, neuro-disability can be for the length of the life of a person and how much that would add to the cost in the healthcare system. So in a way, I already know that, you know, this might be a cost-effectiveness if it can prevent some disability, although we're still working on that answer, but how expensive are these monitors? Is this available to all institutions or do you really need to be a really sophisticated tertiary or portinary hospital to have access to these modalities? Yeah, I think unfortunately the cost and the infrastructural framework is the biggest limiting factor at this point for wide scale scalability. Like I'll give you an example. I consider myself a very passionate advocate for, you know, non-invasive neuro-ultrasound, which is actually one of the cheaper devices. And I have had infrastructural challenges of just integrating blood pressure and TCD values into one monitor. And we are still struggling with that. And it's in the, even the basic infrastructure costs in the range of five figures. And then you add the hidden costs of, you know, people gathering the data, people looking at the data. So cost has been the biggest challenge. And again, you can't put a cost on a dollar value on human life and neurological disability, but I think you still, somebody has to pay for things to move. To add to that, I think, you know, so that's, you know, invasive monitors, you know, but I think with some non and other additional non-invasive monitoring, including both TCDs and also EEG, technician, technologist support. So, you know, it's not just the device and the equipment as well, but the manpower that sort of, that comes with it. And if you want this to be available at all times and obviously having the resources and the manpower available 24 seven as well, that goes along with these devices and monitoring methods. Yeah. I wonder, you know, as we come down to the last few minutes of our discussion, I want to ask you both this question and I have a feeling I might know the answer, but I want to, I'm going to ask it anyway. It's a two-part question. The first part is if you had a magic wand to fix something about neuromonitoring, something that you think could really help us understand brain injury, what might that be? And second part of that question is if you can only have one neuromonitoring modality, what might that be? I'll start with you, Dr. Safar. Oh, with me? So the first question was, what is it that we can do to improve? Yeah. If you had like a magic wand that you can fix something, you know, just instantly about what we know or we don't know about the neuromonitoring that we currently use, you know, something that you always want to be like, oh, this is an irritating, frustrating part of neuromonitoring. What might that be? And the second part is if you can only have one modality, what might that be? So I think it's hard. The first question is, both questions are very hard, but I know we're running short of time. So I think the first thing that I'd like to improve or, you know, sort of really get a handle on for me really is trying to design the right trial. That's, I think, the key thing and how to integrate all of this information into a single trial to do a robust comparative effectiveness study to get answers. And then I think if there was one modality that I would pick, I'm an EEG-er. If I have to, I would say EEG, but I think in general, my general approach is it's impossible to stick to one modality. It has to be a composite and it's not a, you know, one hat fits all. It depends on what is going on with the patient. Thank you. Thank you so much for that answer. How about you, Dr. Somwal, what are your thoughts? I don't think I'm as sophisticated as Sahar, so I'm going to give cheesy answers. I think if I could change one, if I had a magic wand, magical answers, if I had one Disney moment on what I could change with neuromonitoring infrastructure is get the Apple team to work on neuromonitoring and put it into a device that has multiple modalities and puts it in affordable cost. I think that would be one thing that would definitely change this field. If I personally had to choose one pick, you know, obviously I'm a neuroultrasound enthusiast, but on the magic wand side, if I could pick one thing was that I could monitor on a patient would be read a patient's mind. I think if there was a way I could interact with the patient and get their neuro exam, even for the patient with neurological disabilities, you know, new exam is going to be the eventual monitor and no matter what we do. So those are my cheesy answers. It actually ended up being a very insightful answer, not too cheesy, but I do want to thank the two of you. This has been really a fascinating discussion. I'm walking away with a new understanding and a new appreciation for neuromonitoring. And I think the future is young. I mean, the field is young, but the future is bright with researchers and thoughtful clinicians like you guys in the field asking these questions. Thank you again for your time and enjoy the rest of the society meeting. And thanks again for the work that you do with our neurologically injured patients. Thank you. Thank you so much for having us. This was wonderful. Thank you guys. Thanks.

multimodal neuromonitoring

outside of the ICU

EEG monitoring

transcranial dopplers

brain tissue oxygenation monitors

non-invasive measures

pupillometry

intracranial pressure monitors