WEBVTT 00:00:03.497 --> 00:00:10.497 This is the Convergent Science Network podcast. Leading researchers in the domain 00:00:10.497 --> 00:00:16.777 of neuroscience, brain theory and technology are interviewed by Paul Verscher and Tony Prescott. 00:00:22.457 --> 00:00:31.097 So it's the Convergent Science Network podcast at our 10th anniversary BCBT summer school. 00:00:31.217 --> 00:00:36.297 I'm here with Aurore Thibault of the University of Liège in Belgium and University 00:00:36.297 --> 00:00:39.517 Hospital where she investigates consciousness. 00:00:40.817 --> 00:00:47.337 And in your talk this morning, you gave us a very extensive overview of the 00:00:47.337 --> 00:00:53.397 different approaches that people take to assess clinically the level of consciousness of patients. 00:00:53.397 --> 00:01:00.277 So, what are the exact problems there that you face in the clinic in assessing consciousness? 00:01:00.277 --> 00:01:03.477 Consciousness so um we have there are 00:01:03.477 --> 00:01:08.117 different uh types of problems that we can face so we 00:01:08.117 --> 00:01:11.717 have to understand that those patients they have really 00:01:11.717 --> 00:01:14.837 really severe brain injuries that we can't always 00:01:14.837 --> 00:01:18.997 um objective and so some 00:01:18.997 --> 00:01:22.177 patients they have aphasia so they can be fully conscious but 00:01:22.177 --> 00:01:25.077 they just don't understand when you are talking to them 00:01:25.077 --> 00:01:27.777 or um you they can 00:01:27.777 --> 00:01:31.897 have severe motor or impairment or such 00:01:31.897 --> 00:01:34.857 a lot of pain that also reduce the level of consciousness 00:01:34.857 --> 00:01:38.217 or like the way they can express consciousness so we 00:01:38.217 --> 00:01:41.437 have um we have to understand those patients 00:01:41.437 --> 00:01:47.217 we are we need ourselves to be patients so this is from the clinical point of 00:01:47.217 --> 00:01:52.857 view that um that it doesn't mean that if a patient when we when we ask something 00:01:52.857 --> 00:01:58.737 if we don't see something it's not that the patient is not able to do it or doesn't want to do it, 00:01:58.837 --> 00:02:03.937 but it may be because his state of vigilance is low or he's in pain or he's 00:02:03.937 --> 00:02:07.977 just paraplegic, such things. So this is from the... 00:02:09.959 --> 00:02:14.439 The clinical point of view, like from a clinician. But we also have to face 00:02:14.439 --> 00:02:16.959 families' expectations, for instance. 00:02:17.479 --> 00:02:21.139 So we need to, of course, we want to see a sign of consciousness. 00:02:21.659 --> 00:02:26.799 We saw earlier this morning that it's so important to detect such signs of consciousness 00:02:26.799 --> 00:02:33.139 because we know that it will affect the prognosis, but also it may affect end-of-life decision and. 00:02:33.719 --> 00:02:40.319 A rehabilitation program, the effort that will be put to bring these patients 00:02:40.319 --> 00:02:42.519 to a good rehabilitation program and so on. 00:02:42.739 --> 00:02:45.879 So it's really important to detect this first sign of consciousness. 00:02:46.079 --> 00:02:50.379 But on the other hand, we need to be clear and we need to be sure when we see 00:02:50.379 --> 00:02:57.019 something that it is true and it's something that is conscious and not just a reflex. 00:02:57.279 --> 00:03:00.599 So somehow it's really hard that we have this scale. And so, 00:03:00.619 --> 00:03:04.599 for instance, we have the auditory function and the common following. 00:03:04.819 --> 00:03:08.899 And the patient needs to do it like at least three times out of four. 00:03:08.939 --> 00:03:13.099 So I'm going to ask you, try to squeeze my hand really, really strongly. 00:03:13.419 --> 00:03:17.479 And if the patient does it only twice out of four, then it doesn't count. 00:03:17.779 --> 00:03:21.939 And sometimes also it's because that's the rule. We need to have rules and we 00:03:21.939 --> 00:03:24.819 need to be sure that it's not like a grasping reflex, for instance. 00:03:24.819 --> 00:03:29.319 So this for us it's also hard, 00:03:29.399 --> 00:03:33.259 but for the family it's really hard to make them understand that all the movements 00:03:33.259 --> 00:03:37.979 they see or sometimes they have the impression that their son they are following 00:03:37.979 --> 00:03:41.599 them in the room and to make them understand that this might not be a sign of 00:03:41.599 --> 00:03:47.079 that it's not a real visual pursuit or such thing so it's really hard sometimes to. 00:03:48.984 --> 00:03:53.124 To disentangle what is actually conscious and what is not. 00:03:54.264 --> 00:03:57.944 And us as clinicians trying to find a sign of consciousness, 00:03:58.124 --> 00:04:02.544 but we need to be 100% sure that it's conscious, and dealing with the families 00:04:02.544 --> 00:04:04.004 with all their expectations. 00:04:05.044 --> 00:04:11.104 So it's sometimes really challenging. So how many different levels or forms 00:04:11.104 --> 00:04:13.424 of consciousness do you distinguish in the clinic? 00:04:13.844 --> 00:04:16.944 So as disorder of consciousness itself, itself so 00:04:16.944 --> 00:04:19.724 far we i mean we really like to 00:04:19.724 --> 00:04:24.024 categorize and to subcategorize um 00:04:24.024 --> 00:04:27.244 so so far consciousness at 00:04:27.244 --> 00:04:30.084 least for patients like us has been seen 00:04:30.084 --> 00:04:33.464 as dichotomic so you are conscious or non-conscious 00:04:33.464 --> 00:04:36.484 now we think that it's maybe more like a continuum 00:04:36.484 --> 00:04:40.204 but now we have different um states 00:04:40.204 --> 00:04:44.644 so we have the patients that are in coma uh no 00:04:44.644 --> 00:04:47.504 eyes opening and no consciousness of their 00:04:47.504 --> 00:04:50.284 environment and then give you the vegetative state 00:04:50.284 --> 00:04:53.684 we prefer now unresponsive wakefulness syndrome for 00:04:53.684 --> 00:04:56.504 the patients uh recovered eyes opening but that 00:04:56.504 --> 00:05:00.104 doesn't mean that they are conscious it's just eye opening and 00:05:00.104 --> 00:05:03.404 it's not even related to sleep wake cycle um your 00:05:03.404 --> 00:05:06.824 physiological sleep wake cycle and then you 00:05:06.824 --> 00:05:09.764 have this minimally conscious state where the patients are 00:05:09.764 --> 00:05:12.784 minimally conscious but they are still 00:05:12.784 --> 00:05:18.224 not able to communicate so you can't interact functionally with them but they 00:05:18.224 --> 00:05:24.544 can show some sign of consciousness and in this state there are the minimally 00:05:24.544 --> 00:05:27.984 conscious state minus and non-reflexive movement and then when they recover 00:05:27.984 --> 00:05:30.984 some language abilities such as common following, 00:05:31.264 --> 00:05:34.584 intelligible verbalization or intentional communication. 00:05:35.024 --> 00:05:39.124 We say that they are in minimally conscious state, plus because they can understand. 00:05:40.124 --> 00:05:42.924 Us a bit. So now, 00:05:43.862 --> 00:05:46.842 How many patients are we talking about? How many? 00:05:47.022 --> 00:05:53.742 Patients. So in the general population, how many patients do you encounter? 00:05:54.022 --> 00:05:57.202 So luckily, it's just a really very small population. 00:05:58.182 --> 00:06:03.542 So for instance, it's a couple of hundreds I know in Belgium and same in the Netherlands. 00:06:03.802 --> 00:06:11.782 So it's really not a lot. But that's also something that's why it's so hard 00:06:11.782 --> 00:06:15.222 to find treatments for them because it's not a stroke or Alzheimer's. 00:06:15.322 --> 00:06:21.422 We have so many pharmaceutical companies that will try to sponsor your study. 00:06:21.542 --> 00:06:26.782 So it's really hard to get attention for such a small population of patients. 00:06:27.342 --> 00:06:33.742 And then what's the accuracy we have today in assessing the state of these patients? 00:06:34.422 --> 00:06:40.442 So, clinically, we know that if we use the GCS or the Glasgow Coma Scale, 00:06:40.582 --> 00:06:44.142 which is the one that is the most used in intensive care, where we really need 00:06:44.142 --> 00:06:46.142 to assess correctly those patients in fact, 00:06:46.322 --> 00:06:51.122 the rate of clinical misdiagnosis is still 30%, so it's really high. 00:06:51.562 --> 00:06:56.962 Then if we use a more accurate scale, such as a Coma-Virgil scale revise, 00:06:57.442 --> 00:07:06.422 so yeah, true, Maybe we, clinically, we will be able to correctly assess, like, we think, 00:07:06.582 --> 00:07:11.722 we thought before, 100% of the patients, because it is still the best clinical 00:07:11.722 --> 00:07:12.902 scale, the most accurate. 00:07:13.062 --> 00:07:18.662 But still, even though those patients are clinically unconscious, 00:07:19.042 --> 00:07:24.982 when we assess the patient with different types of neuroimaging tools, 00:07:25.282 --> 00:07:33.302 then we still see that about 30% of So one third present brain activity that 00:07:33.302 --> 00:07:37.922 is closer to either minimally conscious state or even higher state of consciousness. 00:07:38.922 --> 00:07:43.762 Okay, but then you also in your talk described a new approach, 00:07:43.922 --> 00:07:52.002 which was combining perturbations of the brain using TMS with different complexity 00:07:52.002 --> 00:07:56.202 measures to look at what the impact is of these perturbations on brain networks. 00:07:56.202 --> 00:08:00.382 Works, and you show to us that if you use the appropriate classifiers, 00:08:00.442 --> 00:08:04.722 like support vector machines, I think is what you used, that you can then rather 00:08:04.722 --> 00:08:07.842 accurately distinguish these different patient groups. 00:08:08.502 --> 00:08:13.702 So do you really see that then as a solution to this diagnostic problem you're describing? 00:08:16.631 --> 00:08:22.171 I think in the intensive care or like early rehabilitation units, 00:08:22.211 --> 00:08:25.951 it's still not a solution because I think it's not, 00:08:26.011 --> 00:08:35.791 it's in too early stage and it's almost impossible in this stage to implement that in clinics. 00:08:35.971 --> 00:08:39.011 It's really made for research centers. 00:08:39.911 --> 00:08:46.151 So in that like practical point of view, it's, we will not be able to do that today or tomorrow. 00:08:46.631 --> 00:08:51.031 But I think maybe in the near future, we are working on that on different paradigms 00:08:51.031 --> 00:08:59.291 or like using other stimuli rather than this TMS, this transgenic magnetic stimulation, 00:09:00.411 --> 00:09:10.071 and also simpler or easier to interpret or direct like an algorithm that could 00:09:10.071 --> 00:09:13.891 give direct feedback of the state of the patients like this PCI, 00:09:13.911 --> 00:09:17.431 this perturbational complexity index. If this could be done. 00:09:18.791 --> 00:09:22.631 If we're going to get a response like two minutes after, then yes, 00:09:22.851 --> 00:09:29.491 I think at this point, this could be useful, even though we are still investigating 00:09:29.491 --> 00:09:33.391 in that technique. And we discussed that earlier today. 00:09:33.631 --> 00:09:41.311 We know that we said that so far, the accuracy is 100% and that we can detect 00:09:41.311 --> 00:09:43.391 consciousness at the single patient's level, 00:09:43.531 --> 00:09:49.691 which is already a huge step as compared to other techniques such as MRI and PET scan. 00:09:49.851 --> 00:09:55.071 However, that doesn't mean we still need to understand that if a patient is 00:09:55.071 --> 00:09:59.591 diagnosed as being in a vegetative state because of the index, 00:09:59.591 --> 00:10:04.051 That doesn't mean that for sure he will never recover because we did that with 00:10:04.051 --> 00:10:06.231 like subacutocronic patients. 00:10:06.471 --> 00:10:10.431 And then there were like two years post-injury and then they in vegetative state 00:10:10.431 --> 00:10:11.931 or unresponsive bifurcated syndrome. 00:10:12.271 --> 00:10:17.671 And then we did this TMS EEG and we saw that they were above this threshold. 00:10:17.751 --> 00:10:21.151 And then, okay, and one year forward, indeed, they recovered. 00:10:21.471 --> 00:10:25.411 That's great. But we now have to do that like at different time points. 00:10:25.411 --> 00:10:29.871 At one week post-injury, at one month post-injury, and so on. 00:10:30.031 --> 00:10:34.171 And just to have a better understanding of when can it predict the recovery. 00:10:34.351 --> 00:10:38.131 If it's only for chronic patients, or it can be also for acute or subacute. 00:10:39.251 --> 00:10:44.331 And this still we don't know. So we still have to investigate in that sense a lot. 00:10:45.851 --> 00:10:50.491 So if you look at this classification, you've also distinct threshold values, right? 00:10:50.571 --> 00:10:54.211 Where you say, well, we lower this threshold of complexity in the response. 00:10:54.211 --> 00:10:55.391 It's a vegetative state. 00:10:55.711 --> 00:10:59.131 If it's above that threshold, it's minimally conscious. 00:10:59.651 --> 00:11:02.351 And then there's a next threshold after which you might say, 00:11:02.431 --> 00:11:06.251 well, this patient is actually normal, probably asleep or recovered. 00:11:07.391 --> 00:11:12.671 So, but now if you talk about life and death decisions, right? 00:11:13.131 --> 00:11:17.851 Do you really see this measure that there is going to be a threshold value where 00:11:17.851 --> 00:11:22.891 you're going to say, well, below this threshold, we can switch off the machines? 00:11:24.643 --> 00:11:31.103 At this point, we can't do that. Sure. We can't because, as I said, we don't know when. 00:11:32.183 --> 00:11:35.283 I mean, we did that again with chronic patients, and then they recovered. 00:11:35.543 --> 00:11:39.903 But now in the acute stage, we can't use that threshold because we know that 00:11:39.903 --> 00:11:43.723 the brain, maybe at that time, we were sure that the patient were in unresponsive 00:11:43.723 --> 00:11:46.783 wakefulness syndrome, but that doesn't mean that, I mean, this patient that 00:11:46.783 --> 00:11:48.783 had the injury three days before, 00:11:49.103 --> 00:11:51.243 that maybe two days later is going to recover. 00:11:51.403 --> 00:11:54.443 That we don't know. So today, we really can't use that. 00:11:54.643 --> 00:11:57.623 Technique but do you think it will go in that direction 00:11:57.623 --> 00:12:01.283 would you recommend it i think 00:12:01.283 --> 00:12:04.303 we need to to do studies to yeah to do 00:12:04.303 --> 00:12:11.423 like um tds sorry tms eg in the intensive care like very early as we do the 00:12:11.423 --> 00:12:18.483 n20 the somatic sensory potential just to and so far this is the only technique 00:12:18.483 --> 00:12:23.383 and only diagnostic too that we can use to make like and with like a very high, 00:12:24.403 --> 00:12:29.663 sensitivity rate to make sure the patient is the chance of recovery are like 00:12:29.663 --> 00:12:35.103 minimal and so and we know that this kind of like. 00:12:36.123 --> 00:12:41.843 Treatment withdrawal are done based on this NGT test but that's the only one 00:12:41.843 --> 00:12:48.323 so if one day TMS-CG is gonna be used for that maybe but but I'm not sure. 00:12:48.583 --> 00:12:52.543 I think it's going to be a good predictor of recovery, but I'm not sure if we're 00:12:52.543 --> 00:12:55.683 going to be able to use it as early as the intensive care. 00:12:56.503 --> 00:13:01.943 Okay. So then, but how objective is the method really? 00:13:02.383 --> 00:13:07.503 Because, so you depend now on the TMS, that means you have to stimulate at a 00:13:07.503 --> 00:13:10.343 certain location with a certain intensity and a certain duration. 00:13:10.603 --> 00:13:15.283 You have to measure at other locations, But now, dependent on the lesions that 00:13:15.283 --> 00:13:19.543 patients have, you might have to induce variability because an area where you 00:13:19.543 --> 00:13:22.743 stimulated one patient, that area isn't there anymore in the other patient. 00:13:22.863 --> 00:13:26.803 So how do you assure that the method stays objective? 00:13:27.983 --> 00:13:32.163 So far, I think that there are really a few patients where we couldn't do it. 00:13:32.223 --> 00:13:39.183 That was impossible to apply to tests because we have different stimulating areas. 00:13:39.183 --> 00:13:41.963 Area and so so far we're always able 00:13:41.963 --> 00:13:44.783 to to to find one that where we could like we were 00:13:44.783 --> 00:13:48.383 sure of what we were recording was true um but 00:13:48.383 --> 00:13:52.323 that's true we have to adapt to the patients uh also something that is really 00:13:52.323 --> 00:13:59.143 important that like now it takes hours to do it so that means that we can't 00:13:59.143 --> 00:14:06.183 in the if we take the example of the intensive care we can't uh apply this tms for 00:14:06.423 --> 00:14:10.183 two hours, it's impossible. It's like practically impossible. 00:14:10.583 --> 00:14:16.063 So there are like many, many limitations that are... That's why it's still in the research process. 00:14:16.163 --> 00:14:20.723 It's not going to be tomorrow that we are going to use it in clinic. But... 00:14:22.123 --> 00:14:26.983 But yeah, there are a lot of limitations that we need to surpass and to see 00:14:26.983 --> 00:14:31.603 how accurate it can be, where we have to place the stimulation, 00:14:32.483 --> 00:14:34.683 depending on the patient's brain lesion. 00:14:34.883 --> 00:14:39.583 Or if we can, a solution, and this is what we are going to do soon, 00:14:39.643 --> 00:14:43.403 is to test other types of stimulation. 00:14:43.623 --> 00:14:52.163 Like, for instance, maybe we can use sounds, which is way less invasive than TMS. 00:14:52.323 --> 00:14:57.263 But as you said, like TMS, the problem is if we are stimulating a brain vision 00:14:57.263 --> 00:15:02.163 that is dead, we don't watch if the patient is dead. So we are always going to face limitations. 00:15:02.563 --> 00:15:10.143 But I think the thing is we need to see what could be used in most of the patients 00:15:10.143 --> 00:15:13.583 and what can be easily implemented into clinics. 00:15:13.883 --> 00:15:16.943 Because research is beautiful, it's really exciting, but at 00:15:16.943 --> 00:15:19.803 the end the aim is to be able to use that technique 00:15:19.803 --> 00:15:22.723 to have like a proper diagnostic so for 00:15:22.723 --> 00:15:28.023 example in yes we are really lucky so we can see the patients for a week and 00:15:28.023 --> 00:15:32.483 then we have we can do tms eeg we can do fmri we can do pet scan we can do high 00:15:32.483 --> 00:15:37.163 density eegs and then with all those tests and like many many clinical assessments 00:15:37.163 --> 00:15:41.423 and we can pose a diagnostic and even though we're not like 100% sure, 00:15:41.603 --> 00:15:45.003 but if it's only high density, she showed some sign of consciousness. 00:15:45.403 --> 00:15:50.423 So it's always a long discussion, but even with all those techniques that are 00:15:50.423 --> 00:15:54.423 amazing, and I'm not sure how many centers can provide that, 00:15:54.623 --> 00:15:58.163 we're still not 100% sure sometimes. 00:15:58.383 --> 00:16:01.183 We receive some dose and then we follow the patient and you can see, 00:16:01.223 --> 00:16:02.983 oh yes, this patient recovers something. 00:16:02.983 --> 00:16:06.083 Thing and so yeah it's 00:16:06.083 --> 00:16:10.923 still a discussion but we need to make neuroimaging tools 00:16:10.923 --> 00:16:13.663 available and that can 00:16:13.663 --> 00:16:16.403 be used in in clinics so that's why i think 00:16:16.403 --> 00:16:19.563 that eeg is a really good options we can find like 00:16:19.563 --> 00:16:23.983 a bit entropy like it's like just three electrodes and we can already do good 00:16:23.983 --> 00:16:29.083 measurements but we also need to find the algorithm behind just to make them 00:16:29.083 --> 00:16:35.803 like to so they can analyze the data like real time and so can give a feedback, 00:16:36.103 --> 00:16:37.703 like a real-time feedback. 00:16:37.863 --> 00:16:43.143 So we talked about the communication in the fMRI and then, but what... 00:16:44.296 --> 00:16:48.256 I mean, it's amazing to know that 30% of the patients are actually conscious, 00:16:48.356 --> 00:16:53.016 even though they can't show that to us and to their family. 00:16:53.216 --> 00:16:59.296 But we can't bring the patients to the fMRI every time we want to ask them a question, right? 00:16:59.556 --> 00:17:04.036 Right. But then, so, again, I want to have, of course, the clinical objectives 00:17:04.036 --> 00:17:07.336 of these measures, which are really very important and relevant. 00:17:07.596 --> 00:17:12.316 But on the other hand, studying these patients has also helped us to understand 00:17:12.316 --> 00:17:17.136 consciousness. And this is also one reason why this whole research field is 00:17:17.136 --> 00:17:19.396 driving itself forward. 00:17:20.376 --> 00:17:25.136 In your presentation, you made an important distinction there between what you 00:17:25.136 --> 00:17:28.396 called an internal consciousness and an external consciousness that you felt, 00:17:28.516 --> 00:17:33.076 at least in these patient groups you look at, is always popping out as an important 00:17:33.076 --> 00:17:34.016 distinguishing feature. 00:17:34.436 --> 00:17:38.176 So what do you mean exactly with internal and external consciousness? 00:17:38.176 --> 00:17:44.076 Consciousness oh so um so the internal consciousness is the consciousness of 00:17:44.076 --> 00:17:51.096 yourself so what we call also like the little voice okay so when you are like oh for instance you you, 00:17:51.736 --> 00:17:55.436 hurt yourself and you're in pain and then you focus on yourself so this is like 00:17:55.436 --> 00:17:59.316 really or when you your internal thoughts that's the internal consciousness 00:17:59.316 --> 00:18:03.556 and then external consciousness gonna be the consciousness of your environment 00:18:03.556 --> 00:18:07.816 so if you are listening to someone talking This is the external consciousness. 00:18:08.296 --> 00:18:12.576 And they are anti-correlated. That doesn't mean that if one is up, the other one is zero. 00:18:12.816 --> 00:18:17.976 But that's usually if you're like fully concentrated on yourself, 00:18:18.176 --> 00:18:20.576 you will be less aware of what is happening around you. 00:18:20.776 --> 00:18:24.756 And this has been shown many times and the contrary as well. 00:18:24.976 --> 00:18:30.876 And so we did a study where we compared those two kinds of consciousness. 00:18:30.876 --> 00:18:33.716 So more internal related issues. 00:18:36.158 --> 00:18:39.078 Thought and then more external related thought and then 00:18:39.078 --> 00:18:42.098 we compared the the network that were activated for 00:18:42.098 --> 00:18:44.858 one and the other and so we saw that we had 00:18:44.858 --> 00:18:47.998 like the external consciousness network which is the lateral 00:18:47.998 --> 00:18:50.678 frontal parietal cortices and then the 00:18:50.678 --> 00:18:55.978 internal one uh which is more like the mesofrontal cortex anterior singular 00:18:55.978 --> 00:19:02.538 cortex and then precuneus and and um posterior singular cortex so one is external 00:19:02.538 --> 00:19:07.298 for the external consciousness network and one is more internal for the internal 00:19:07.298 --> 00:19:08.098 consciousness network. 00:19:08.278 --> 00:19:11.638 And normally they should be anti-correlated. 00:19:11.838 --> 00:19:14.058 So when one is activated, the other one is not. 00:19:14.638 --> 00:19:19.938 So the internal consciousness system, you would then equate with a more metacognitive 00:19:19.938 --> 00:19:21.238 self-reflective state. 00:19:21.818 --> 00:19:26.218 It's not necessarily experience of hunger or of a state of the body. 00:19:26.278 --> 00:19:29.038 It's really more a metacognitive self-reflective state. 00:19:29.318 --> 00:19:32.898 Yeah. This is how we do it. But maybe we do it like, you know, 00:19:32.918 --> 00:19:39.798 more simplistic way of the way we categorize consciousness between two components, 00:19:40.118 --> 00:19:42.538 awareness, wakefulness, but yes. 00:19:43.218 --> 00:19:49.858 But then do you envision that also as, let's say, two subsystems of consciousness that sort of. 00:19:51.633 --> 00:19:54.653 Exclusively are active it's either one or 00:19:54.653 --> 00:19:57.513 the other no it's not exclusive okay it's not exclusive 00:19:57.513 --> 00:20:00.353 i mean i i'm not saying that again if you are listening 00:20:00.353 --> 00:20:03.593 to me right now you are not you're not 00:20:03.593 --> 00:20:09.013 conscious of yourself at all it's like but when one is highly activated then 00:20:09.013 --> 00:20:13.233 the other one is less activated so they are in a competitive relation is that 00:20:13.233 --> 00:20:17.613 what you would say are they competing i'm not i'm not sure i would say that 00:20:17.613 --> 00:20:22.033 they are competing but at least for instance what But with a new study, 00:20:22.193 --> 00:20:25.193 there we were focusing on the default mode network. 00:20:25.653 --> 00:20:35.093 And so we know that within the network, the area needs to be highly connected. 00:20:35.293 --> 00:20:38.573 But then we found out that in patients with disorder of consciousness, 00:20:38.813 --> 00:20:43.393 we also have hyper-connectivity, which is pathologic. 00:20:43.393 --> 00:20:50.813 So, for instance, in LC control, when the default mode network is highly connected, 00:20:50.953 --> 00:20:54.113 then there is no connection with other brain regions. 00:20:54.113 --> 00:21:02.693 While in patients, you have this hyper-connectivity, meaning that concurrently, at the same time. 00:21:03.273 --> 00:21:07.753 The areas that should be connected in the default mode network, 00:21:07.953 --> 00:21:11.193 then other areas are also activated. 00:21:11.453 --> 00:21:15.693 And this is like the pathological hyper-connectivity only seen in patients with 00:21:15.693 --> 00:21:18.693 disorder of consciousness. But in this case, connectivity would mean some correlation 00:21:18.693 --> 00:21:24.593 structure in the data that you measure from the brain, either EEG or fMRI. So here's an fMRI. 00:21:25.273 --> 00:21:29.813 So it's a rather slow signal, and it's related to, let's say, 00:21:29.873 --> 00:21:33.273 correlations among the different measurement points. 00:21:33.493 --> 00:21:40.413 But now, would that mean that some correlations go up that connect these subsystems? 00:21:40.433 --> 00:21:44.853 Or are the intrinsic correlations within the subsystems going down, 00:21:44.953 --> 00:21:51.593 so you are revealing the inter-subsystem correlation? Can you distinguish these two explanations? 00:21:55.009 --> 00:22:03.529 Those are really good questions, but do you mean like, I would say that depending 00:22:03.529 --> 00:22:08.389 on the state of your mind, 00:22:08.549 --> 00:22:11.169 then one system will be first activated, 00:22:11.369 --> 00:22:13.649 and then the other one will be less. 00:22:14.509 --> 00:22:19.469 From a pure, let's say, fMRI perspective, what I'm measuring is correlations 00:22:19.469 --> 00:22:23.969 among my voxels, right? And you could say for the control condition, 00:22:24.369 --> 00:22:29.229 I see a certain correlation within these two subsystems, and I don't see it between. 00:22:30.069 --> 00:22:34.129 But now you could argue that underlying that is some causal structure, 00:22:34.289 --> 00:22:39.109 and that maintains this intra-module correlation. 00:22:40.229 --> 00:22:43.269 And as a result, you don't see any inter-module correlation. 00:22:43.809 --> 00:22:48.989 But now in the coma patient, it's this intra-module correlation that you might 00:22:48.989 --> 00:22:54.889 lose. And as a result, you are unmasking this relationship between the two modules. 00:22:55.389 --> 00:23:01.209 So is it really indicative of anything that is specific to these states of consciousness? 00:23:01.329 --> 00:23:06.729 Or is it just like an artifact of, let's say, losing internal coupling within systems? 00:23:06.849 --> 00:23:10.669 So that's a bit the question I'm after. Can we distinguish that? 00:23:10.949 --> 00:23:17.649 I think that in the state that we are now, we are not able to do that as specific. 00:23:17.649 --> 00:23:24.869 Fake but yeah I think it's more like in coma patients then it's going to be 00:23:24.869 --> 00:23:32.349 this more like anarchic activity and as like so many. 00:23:34.376 --> 00:23:37.136 Brain regions are damaged then it can be 00:23:37.136 --> 00:23:39.816 possible that one of them is 00:23:39.816 --> 00:23:42.516 trying to to reactivate it instead of the other one and 00:23:42.516 --> 00:23:47.876 then it just creates some kind of arnotchic uh connectivity that is like in 00:23:47.876 --> 00:23:52.276 fact more that creates more damage that or at least yeah functional damage that 00:23:52.276 --> 00:23:59.656 it if it yeah so more like an arnotchic right so now then um what you also then 00:23:59.656 --> 00:24:01.416 discussed in in that context, 00:24:01.556 --> 00:24:05.276 was the subjective experience in that sense of these patients. 00:24:05.456 --> 00:24:10.136 And in particular, you talked about locked-in syndrome patients where you looked 00:24:10.136 --> 00:24:14.736 at their reported state of, let's say, satisfaction with life or happiness. 00:24:14.836 --> 00:24:20.496 And you had this surprising result that actually they showed levels of satisfaction 00:24:20.496 --> 00:24:23.856 or happiness with life that was comparable to healthy controls, 00:24:24.096 --> 00:24:29.616 which, of course, to us sounds very counterintuitive because their bandwidth 00:24:29.616 --> 00:24:31.976 of communication with the world is severely reduced. 00:24:33.436 --> 00:24:35.616 So how do you interpret that result? 00:24:35.796 --> 00:24:40.036 Do you really think it is an accurate reflection of their state of being? 00:24:41.156 --> 00:24:47.816 Or is there some, let's say, recalibration also for them of what happiness means? 00:24:48.556 --> 00:24:52.376 I think the two propositions are correct. 00:24:52.836 --> 00:24:58.496 So first, it needs to be said, and it was acknowledged in the paper, that, 00:24:59.825 --> 00:25:06.245 We send the questionnaires to thousands of patients and only a small proportion replied. 00:25:06.605 --> 00:25:11.885 So it's possible that we got the reply from people that are actually not that 00:25:11.885 --> 00:25:13.445 depressed. That's a possibility. 00:25:13.965 --> 00:25:21.105 So maybe this really interesting result where we see that the patients in Lactin 00:25:21.105 --> 00:25:28.265 syndrome on average are not unhappy, but they looked as happy as LC controls. 00:25:30.265 --> 00:25:34.285 I mean, on average, again, I mean, not at a single subject level, 00:25:34.465 --> 00:25:36.805 but so that's a possibility. 00:25:37.045 --> 00:25:41.345 And then that's true to that those patients, 00:25:41.405 --> 00:25:48.225 they can find the happiness in other things that we first is almost impossible 00:25:48.225 --> 00:25:52.825 to conceive because we are not in, I mean, the way they are. 00:25:52.825 --> 00:25:58.285 I met a lot of locked-in patients, and they are going on vacation. 00:25:58.605 --> 00:26:01.405 That's why it's like, oh, how can they go on vacation? But they do, 00:26:01.505 --> 00:26:04.105 and they go like twice, three times a year. 00:26:04.385 --> 00:26:08.125 And they still have like a really good relationship with their family. 00:26:08.565 --> 00:26:15.005 But one thing that we found out in that study and other follow-up studies is 00:26:15.005 --> 00:26:18.105 that the more a patient is able to interact with the environment, 00:26:18.285 --> 00:26:22.005 the happier he is, which makes sense. So that means that that's true. 00:26:22.125 --> 00:26:28.605 Interaction with your environment is an important factor for someone to be happy. 00:26:29.025 --> 00:26:31.965 So I don't know. 00:26:31.985 --> 00:26:35.225 It would be so interesting, for instance, that we had those patients that are 00:26:35.225 --> 00:26:39.645 in complete locked-in syndrome when even they can't move their eyes. 00:26:39.945 --> 00:26:43.825 So the only way, I mean, we had a case of patients like that. 00:26:43.925 --> 00:26:50.205 So it was impossible to know that the patient was conscious without neuroimaging. 00:26:50.205 --> 00:26:56.845 It was impossible, but we did this, that she was conscious thanks to the PET scan. 00:26:57.305 --> 00:27:01.225 And we thought, oh my God, the PET scan looks like you and me. 00:27:01.305 --> 00:27:01.965 I mean, it was impressive. 00:27:02.405 --> 00:27:07.245 And then we did other tests and they all said, okay, yes, indeed, 00:27:07.425 --> 00:27:10.245 it looks like she's really, she's conscious. 00:27:10.405 --> 00:27:17.445 But still, it was, she couldn't communicate like how we are used to see in Lactin 00:27:17.445 --> 00:27:24.545 syndrome with the eyes no movement at all nothing and so yeah can we say that 00:27:24.545 --> 00:27:27.665 this patient might be happy in a way. 00:27:29.083 --> 00:27:32.183 I have no idea, but this is a question that could be asked. 00:27:32.483 --> 00:27:35.623 But now with these patients, they also showed 00:27:35.623 --> 00:27:44.743 us that there are methods that use fMRI recordings with mental imagery to help 00:27:44.743 --> 00:27:49.843 people to at least give yes or no answers by imagining that they are behaving 00:27:49.843 --> 00:27:52.983 in some context or navigating some environment, right? 00:27:52.983 --> 00:27:58.403 Do you think this would be a useful interface for all these patients, 00:27:58.543 --> 00:28:01.023 or do you think it's too cumbersome and it's not going to work yet? 00:28:01.283 --> 00:28:07.743 I think the EEG one is going to maybe be easier to apply in clinics again. 00:28:08.523 --> 00:28:17.023 But also the thing with all those motor imagery is that we have a really high rate of false negative. 00:28:17.703 --> 00:28:22.203 So many patients in minimally conscious state that are clinically answering 00:28:22.203 --> 00:28:24.923 to comments, they can't perform the task. 00:28:25.383 --> 00:28:30.163 And also 75 to 80% of the LC control, they can't do it. 00:28:30.383 --> 00:28:35.303 So we also need to find a way to make sure like an easy way and in a way like 00:28:35.303 --> 00:28:39.423 a task, a motor test can be, I mean, done by almost everyone. 00:28:40.043 --> 00:28:44.203 And how are you going to deal with the variability of the neural response across all these patients? 00:28:44.983 --> 00:28:49.443 Yeah, repeated assessments, I guess. Mm-hmm. Okay. I mean, there is no other solutions. 00:28:51.283 --> 00:28:55.843 There's a really nice study that was published really recently in Archaeophysiology, and, 00:28:56.923 --> 00:29:02.263 we have to do at least five clinical assessments, so if we take these five clinical 00:29:02.263 --> 00:29:07.263 assessments, maybe that we need to do five FMRI or five EEGs, 00:29:07.423 --> 00:29:09.923 and yeah, it's kind of burdensome, but. 00:29:11.313 --> 00:29:14.113 They're fluctuating so we have no other options we have to 00:29:14.113 --> 00:29:16.933 could you imagine that that in those patients we we 00:29:16.933 --> 00:29:23.333 start to implant devices an e-cock or something to measure uh states of the 00:29:23.333 --> 00:29:27.413 brain to actually help communication would you see that as a way forward yeah 00:29:27.413 --> 00:29:31.993 i think we we i mean not us but other laptop doing it already with locked in 00:29:31.993 --> 00:29:34.473 patients so yeah this is something that may be 00:29:34.513 --> 00:29:37.353 useful and and and maybe you're sorry to find like 00:29:37.353 --> 00:29:40.493 perfect time to to to say 00:29:40.493 --> 00:29:44.073 okay this is the moment when we have to do something with the subject right 00:29:44.073 --> 00:29:47.353 so now in your talk you also showed us a 00:29:47.353 --> 00:29:53.593 possible let's say mechanical interpretation of states of consciousness that's 00:29:53.593 --> 00:29:57.573 focused very much on the thermocortical system going back to to nick schiff 00:29:57.573 --> 00:30:03.373 and and others way to look at It's basically the modulation of the thalamo-cortical 00:30:03.373 --> 00:30:06.073 system that will define these states of consciousness. 00:30:07.073 --> 00:30:11.313 And in that specific model, you also look at additional subcortical loops over 00:30:11.313 --> 00:30:12.893 the basal ganglia and so on, right? 00:30:12.953 --> 00:30:16.973 And it's a bit a sort of model you might also apply to Parkinsonism or so on, right? 00:30:17.013 --> 00:30:21.693 So it's really whether you are switching the thalamus in pathological states 00:30:21.693 --> 00:30:25.233 of low frequency bursting. thing. 00:30:25.773 --> 00:30:30.993 Do you feel that that model is sufficient to understand these patients you look 00:30:30.993 --> 00:30:36.373 at or is it more like really a very first approximation of what we need and what is missing? 00:30:36.853 --> 00:30:43.953 So yeah I think so far it's been a really nice model that can explain. 00:30:45.170 --> 00:30:48.410 Why some treatments work and why others don't. 00:30:48.490 --> 00:30:53.470 The most impressive one is the Zalpidem, of course, but so far it fits. 00:30:53.790 --> 00:30:59.670 And for instance, with TDCS, we have tried TDCS in different cortical areas 00:30:59.670 --> 00:31:02.830 and the prefrontal one is the one that works the best. 00:31:03.710 --> 00:31:07.290 So we were, okay, yeah, it makes totally sense. But true, 00:31:07.510 --> 00:31:10.550 I think it's still like an hypothesis 00:31:10.550 --> 00:31:13.690 processes and and but it's a nice 00:31:13.690 --> 00:31:16.870 direction if we need to to try new treatments to 00:31:16.870 --> 00:31:19.650 to at least have a rationale behind what we 00:31:19.650 --> 00:31:25.490 do and and that could be yeah some of some health if we want to target some 00:31:25.490 --> 00:31:31.510 specific brain regions right which is the model does not really account or include 00:31:31.510 --> 00:31:35.710 the role of neuromodulators for instance right it doesn't explicitly take it 00:31:35.710 --> 00:31:38.430 into account right i I think it doesn't take into account everything. 00:31:38.750 --> 00:31:43.990 It's really specific to... It was like... I mean, the model existed before, 00:31:44.170 --> 00:31:48.950 but it was used for disorder of consciousness due to zolpidem because it was 00:31:48.950 --> 00:31:50.990 really specific to this drug. 00:31:51.250 --> 00:31:57.370 And I think that this might be really useful to explain, 00:31:57.510 --> 00:32:02.450 again, some treatments, but maybe only in anoxic patients because maybe the 00:32:02.450 --> 00:32:07.210 stratum is not injured at all in some traumatic patient. and then why TDCS is working. 00:32:07.510 --> 00:32:11.150 So no, of course, I think it doesn't fit everything, but... 00:32:12.438 --> 00:32:16.418 It works well, and maybe it works for very, very specific patients, 00:32:16.498 --> 00:32:21.058 and then we can fully use it to try to treat them and improve their recovery. 00:32:21.338 --> 00:32:23.838 But, yeah, we still need to understand that better. 00:32:24.238 --> 00:32:30.038 Okay. So, look, you finished your PhD in Liège with Steve Lorais, 00:32:30.818 --> 00:32:33.118 and now you did your postdoc in Boston, 00:32:34.158 --> 00:32:40.318 and now you're moving back to Liège to sort of build your career as a young 00:32:40.318 --> 00:32:41.618 and upcoming scientist. 00:32:43.078 --> 00:32:46.258 But now in your experience in this field, not as a clinician, 00:32:46.578 --> 00:32:52.578 what would be Aurora's law to study consciousness and the mind? 00:32:54.718 --> 00:33:01.858 I think I would like to believe that the unresponsive wakefulness syndrome does 00:33:01.858 --> 00:33:06.978 not exist and we don't have the tool yet to detect consciousness in every patient. 00:33:06.978 --> 00:33:10.658 Patients uh we have seen patients that were inanimate positive 00:33:10.658 --> 00:33:13.498 boyfriend syndrome for years and then one day they 00:33:13.498 --> 00:33:18.758 recovered uh like three years later they could they were able to to talk even 00:33:18.758 --> 00:33:22.378 though they because of severe spasticity and motor impairment they will never 00:33:22.378 --> 00:33:27.958 be able to walk again or be like independent but still so this yeah i think 00:33:27.958 --> 00:33:31.778 i would like to to to believe that 00:33:31.818 --> 00:33:39.718 this is a possibility and to try to find a way so we will be able to find this 00:33:39.718 --> 00:33:45.278 consciousness even in patients that are clinically unresponsive. 00:33:45.998 --> 00:33:51.838 So keep hope. Always keep hope. So five years from now, I'm going to come visit 00:33:51.838 --> 00:33:53.818 you in Liège, whether you like it or not. 00:33:55.378 --> 00:33:59.618 And I'm going to check then whether a prediction you're going to make today 00:33:59.618 --> 00:34:02.258 was falsified or verified. 00:34:02.498 --> 00:34:07.458 So what's the one prediction that you would like to see really tested in this 00:34:07.458 --> 00:34:09.718 five-year framework or time frame? 00:34:12.727 --> 00:34:19.427 So I'm going to be really clinical again, but I think that a lot of treatment 00:34:19.427 --> 00:34:22.267 options are available, but we can't use them. 00:34:22.387 --> 00:34:28.767 And I would like to stimulate the translation of such treatment. 00:34:28.807 --> 00:34:32.687 One of them is the TDCS. For me, it's an amazing tool, and that's why I did 00:34:32.687 --> 00:34:35.907 my PhD at the Neuromodulation Lab with Felipe Fregni. 00:34:36.007 --> 00:34:40.027 I think it is so hard to translate that to clinics, And I mentioned before that 00:34:40.027 --> 00:34:45.887 it was really hard to get attention for these very small populations of patients. 00:34:46.027 --> 00:34:54.427 So my hope is that we can have clinical trials and have this translation to 00:34:54.427 --> 00:34:57.427 clinics and use all the treatments we have. 00:34:59.287 --> 00:35:03.147 We have so many things that we can use for such patients that will increase 00:35:03.147 --> 00:35:07.607 either the comfort or the recovery for sure. and this is something that I will 00:35:07.607 --> 00:35:10.327 always work for. So, my hope. 00:35:10.567 --> 00:35:13.207 All right. All right, Thibault. Thank you very much for this conversation. 00:35:13.467 --> 00:35:14.867 Thank you so much for inviting me. 00:35:15.280 --> 00:35:21.360 Music. 00:35:20.727 --> 00:35:26.287 The CSN Podcast was produced by the Convergent Science Network of Biometrics 00:35:26.287 --> 00:35:32.727 and Biohybrid Systems, a project funded by the European 7th Research Framework Program. 00:35:34.227 --> 00:35:39.547 For more interviews, recorded lectures, or upcoming conferences in the field 00:35:39.547 --> 00:35:45.787 of biometrics and biohybrid systems, go to csnnetwork.eu. 00:35:46.000 --> 00:35:53.840 Music. 00:35:46.107 --> 00:35:47.987 And thank you for listening.