Welcome to the first in a series of interviews that I’ll be conducting for PLOS Neuro as Contributing Editor! My regular blog, at Discover Magazine, is Neuroskeptic.
For this first post, I spoke to Dr Srivas Chennu, first author of the PLOS Computational Biology paper “Spectral Signatures of Reorganised Brain Networks in Disorders of Consciousness” which was published on October 16th and has already been viewed over 12,000 times. The paper has also received extensive media coverage. In their study, Chennu and colleagues used graph theoretical analysis to explore the connectivity of brain activity in individuals with a disorder of consciousness – including those in a persistent vegetative state. This research aims both to understand human consciousness and to predict which patients are likely to recover from their brain injuries.
This striking image is a 3-dimensional representation of the data presented in 2 dimensions in the PLOS Computation Biology paper (Figure 2). It shows weighted connectivity networks derived from EEG data. In the alpha frequency band, healthy brain networks were organized into long-range frontoparietal modules. By contrast, in patients with a disorder of consciousness, the alpha networks consisted of spatially localised, weaker modules.
NS: “First of all could you tell me a bit about your own background?”
SC: I’m a senior research associate at the Department of Clinical Neurosciences and the MRC Cognition and Brain Sciences Unit at the University of Cambridge. I’ve been doing a postdoc for five years or so, before that I did my PhD in the UK and that was in computational neuroscience – my PhD bridged from computer science to computational neuroscience, with some experimental work too. My first degree was in computer science and engineering.
“What led you into this area, disorders of consciousness?”
I did my PhD in attention and conscious perception (e.g.), looking at how what you attend to affects how you perceive things, and how salient information is retained into episodic memory and processed. So I got into this field from the perspective of studying consciousness and that’s still our broader aim.
What I do now is work with data from both healthy adults and patients in different altered states of consciousness – including patients with chronic disorders of consciousness (like in this paper) but also acute cases, for example people who are in the immediate aftermath of a brain injury, coming out of an induced coma. That’s one line of research.
I also look at data from healthy volunteers undergoing sedation: how do drugs change brain networks in terms of the mechanisms that support consciousness.
And the third aspect of the whole thing is sleep: we get healthy people to fall asleep in the lab and we measure brain activity, linking the change in brain networks to responsiveness.
So overall it’s brain behaviour linkage across many different states of awareness.
The only thing we haven’t quite gotten going yet, that we’d like to, is hypnosis. It’s been shown a number of times that, in some people, there are interesting dissociations of consciousness that emerge in that case.
“So what you introduce in the new paper is a new analysis approach drawn from graph theory?”
That’s right. Some of graph theory has been around for centuries, and electroencephalography (EEG) has been around for a long time too, but now with technological advances we can measure EEG with the density required to capture enough information such that, combined with methodological advances in graph theory, we can now merge them to study network properties.
The EEG signal gives you two orthogonal bits of information, one is power the other is phase and this allows you to tell how linked or connected part of the brain are. The most obvious aspect of the EEG signal is power, with the naked eye you can see certain waves such as alpha waves – indeed the first thing that was ever characterized on the EEG by Hans Berger was alpha.
What’s less evident is the phase relations and the phase-amplitude relations. So we’re putting a cognitive mechanistic explanation to those phase relationships, what does it tell us about the brain?
Animal evidence reveals that these phase relationships might serve as a means by which information is transferred from here to there by synchronous firing.
“You found that these networks were disrupted in individuals with a chronic disorder of consciousness?”
That’s right, indeed the alpha networks they lost the information carrying capacity that we see in the healthy adult brain. But a major point of the paper is that these patients actually have shifted connectivity into other frequency bands. They do have this robust connectivity characteristic. That’s interesting.
In fact it’s counterintuitive if you thought “Oh well there’s just nothing going on” in these brains, actually there is. The brain activity is not just noise. There are connected structures there, but they’re in lower frequency bands than in the healthy awake brain, and they have abnormal topology as well.
“In the paper you have two separate groups of DoC patients, vegetative state (VS) and minimally-conscious state (MCS). These are defined in terms of behavioural responses (or the lack of them). But in the future, might it be possible to go towards a more neuroscience-based definition?”
Right. Of these abnormal states of consciousness but also by extension of consciousness itself. Our everyday tests of “is somebody conscious?” It’s very simple. I know you’re conscious because if I ask you so, you concur. That’s not very sophisticated. So we’re going beyond that for the first time – to some extent – because as neuroscience advances the reliability of how we can tell these things, to a place where we can talk about a neuroscience correlate of consciousness.
Because raising your right hand could be a behavioural correlate of consciousness, but it’s not your consciousness itself.
This is a point that Adrian Owen, who as you know has been leading this research, has been saying – using neuroscience as a surrogate. Of course clearly there’s some way to go before this becomes a clinical tool, like a measurement of blood pressure etc. we’re not there yet. But I’m optimistic.
“And do you give individual feedback on the results, for the families and the caregivers of the patients who participate?”
We do, yes. Although we’re not a clinical service, that doesn’t mean that the patient comes in to us, we do a test, get our research data, and then we say “Thank you, goodbye”. We do provide feedback, though with the qualification that it’s experimental not a clinical test, and this is a key part of what we do. We perform a whole series of tests – the patient comes in for a whole week, and we perform a behavioural assessment, eight or nine different MRI tests, and about the same number of EEG tests. We don’t just measure brain networks we also test their vision, hearing, the brain language cortex – a whole bunch of tasks derived directly from cognitive neuroscience, that are already understood and validated in healthy adults.
There’s a hierarchy of tests at different levels of complexity. And all of these provide results, sometimes as an image, and usually with a statistical outcome as well. We can either take the healthy population and ask, how similar is the patient to that. Or in other cases you can do single trial statistical analysis e.g you can say whether a certain function is preserved or not. So these figures and statistics are compiled into a report including behavioural, clinical and imaging measures. This is fed back to the caregiver who consented, and to their neurological care team. And this feedback often makes the families want to come back to us, because they’re given more information than they otherwise would be. We do quite a lot of follow-up assessments, despite the logistical challenges that entails. We hope to see, if a patient recovers and makes clinical progress, the neuroimaging measures should reflect that.
“Have you ever had a patient who’s recovered after taking part in your research?”
Yes, in fact there was a Radio 5 interview the other day (iPlayer from 23:00) with a mother whose son participated in our study and went on to recover. In fact her son started recovering around the time he was in Cambridge participating in our study. He’s made a recovery in that he’s regained consciousness – although he still has speech and motor impediments.
“That’s remarkable. Does he remember taking part in the research?”
That’s exactly what the Radio 5 presenter asked! It’s very interesting – his mother says no, he doesn’t remember anything – it’s a complete blank. And this is a scientific question which is fascinating us. Of course one patient is not a study – but to what extent memory is preserved, well, that actually speaks to a lot of ethical issues for these patients.
Consciousness is not just being aware of something, but also being aware that you were aware of it yesterday. Without continuity of consciousness, the unity of consciousness, is it the same consciousness? And this is relevant to issues such as pain: are you in pain, does a medication help reduce that pain, well how do you know, if you don’t remember what the pain was like yesterday? All these questions are fascinating.
“It sounds like you’re right up against some philosophical issues that otherwise are rather abstract, and here they’re very real.”
Yes – they are very real, and especially when you’re talking to the families whose lives have been upturned by these injuries. And they’re left in a state of tension often, because anything could happen: they might wake up, they might not, and it is currently difficult to know either way. This situation can last for months or years. It’s very hard for the families.
Neuroskeptic is a British neuroscientist who takes a skeptical look at his own field, and beyond. On Twitter @Neuro_Skeptic
The views expressed in this post belong to the author and are not necessarily those of PLOS.