Science and Innovation

Gustavo Deco is Research Professor at the Catalan Institution for Research and Advanced Studies (ICREA) and Professor at Pompeu Fabra University, where he directs the Computational Neuroscience group and the Center for Brain and Cognition. On the occasion of the dialogue organised by the Barcelona Knowledge Hub of the Academia Europaea in which he participated together with Mara Dierssen, from the Centre for Genomic Regulation, we spoke with him about the brain, neuroscience, the advances made and those yet to come.

The brain is one of the most enigmatic things in the world. What do we know about it and what do we still need to know?

The bad news is that we know little about the brain. Despite centuries of research, the number of measurements we have today and the possibilities of characterising the brain, especially over the last century – since Ramón y Cajal, perhaps – we know very little indeed. The good news is that we know why we know little. Because we are dealing with an extremely complex system. And the most important thing is to understand it from a mechanistic point of view. Why it is there and in what way it is performing its functions.

So, the greatest complexity is to understand its mechanism?

Yes. They are trillions of elements, they are complex cells that have a dynamic that can be described mathematically by a differential equation and, by having trillions of elements, their complexity is multiplied many times over. In mathematics, and I’m sorry for getting technical…

No problem.

… represents a system of non-linear coupled differential equations. It’s one of the most complex situations in mathematics. And it’s the reason why it’s so hard for us to understand how the brain performs its operations. Anyway, moving away from the pessimistic message, what really matters is that we know what we have to investigate. And, little by little, we are understanding some essential features underlying that kind of computation. Both in a healthy brain and in a diseased brain.

And what do you think has been the differential advance in recent years?

There are many. It’s difficult to choose. It’s like being asked what your favourite film is, there are too many. But there are advances that subjectively have had an impact on me. One is the mechanistic correlates underlying the different states of the brain. For example, in its states of consciousness: when a brain is awake, when it is sleeping, when it is anaesthetised, when it is in a coma; apart from the fact that we know that the person behaves differently, we also know that the activity of the brain is totally different. Even if the brain is sick.

And what does this knowledge give us?

If instead of asking a person how they feel, I analyse how their brain feels, I will see the reason why that brain is malfunctioning. It gives us an explanation of the disease, even its treatment pathways. We are still at the beginning of this process, but all these routes are opening up thanks to the theoretical framework that allows us to identify mechanisms without relying on observation alone.

Will this change help us understand how to treat a sick brain?

Definitely. Psychiatry still operates in a very empirical way. If you know people who have been to a psychiatrist or therapist, dialogue is the way to objectify what is wrong with them. It’s funny because we know that what’s wrong with them is in the brain and yet we ignore it. If you have a heart problem, the first thing to do is to measure your heart. As well as asking you if you feel agitated, they analyse what’s wrong with the organ. In psychiatry that doesn’t happen yet. Psychiatry is changing radically and I think it is the next revolution. In a matter of a few years, apart from asking the patient what is subjectively happening to him, we are going to tell him that the brain is the origin of his situation.

So neurocomputing can help prevent disease.

Well, I don’t know about preventing. To detect them early and, perhaps, even to design the type of treatment that best suits each person. Individualisation is a big issue in medicine today. Especially in psychiatry, there is no longer a valid taxonomy, there is no disease called schizophrenia or a disease called depression. There is a person who has symptoms and there is a person who has an extremely individual brain condition. So what is more important is to understand how we can help that individual person and not a group of people who have similar symptoms.

And with this personalisation, through what mechanisms would diseases be treated?

The two major lines of treatment are either pharmacological, where you interact at the mechanistic, neuronal and synaptic level, or through electromagnetic stimulation. The big question is whether one can design or predict the effect that a certain pharmacological treatment or a certain electromagnetic treatment will have. I always give a very banal example. In the old TV sets, the ones that worked with valves, when the interference started, we used to get up and with a sharp knock, they  worked. And that’s what we try to do today with stimulation.

And do they work?

Yes, but not always. The system we have, apart from being complex, is extremely robust. However, sometimes an electromagnetic or pharmacological aid puts the system back in place. And today, if we have a blueprint of the brain, instead of just blindly giving it an electromagnetic shock, we can individualise and provide pharmacological or electromagnetic stimulation where we really need to, with a better chance of stabilising it.

We have talked about what is to come. But what technologies do you think have contributed to advances in our understanding of the brain so far this century?

As always in science, the starting point is observation. And there really has been a huge development in the last few decades in brain observation techniques. From the invention of magnetic resonance imaging, with which we can visualise the activity of the brain when it is performing a task. And the second great technological advance is the learning of more refined techniques of computation, of mathematical language. The ways in which we can express and analyse this complexity and, I insist, the computational capacities we have today, which allow us to carry out simulations at the level of the whole brain.

The dialogue at the Academia Europaea’s Barcelona Knowledge Hub, in which you participated together with Mara Dierssen, brought to the table the role of neurocomputing, but also that of neurobiology. What does each of these disciplines contribute to the brain?

If neurocomputing combines concepts and techniques from neuroscience with computer science to study the computational functioning of the human brain, neurobiology is the dimension of experimental observation on a more biological scale, which is absolutely fundamental. Neurobiology allows us to look at the most microscopic parts at the neuronal level and understand how they act when they are performing a certain task. This allows us to obtain complementary information to the theoretical framework, to the computational framework.

In other words, they need each other.

Absolutely, that’s the beauty and also the complexity of neuroscience. Apart from the fact that it is a complex system, it is a multi-scale system. It has different spatial and temporal scales. Neurons function at the level of microseconds, milliseconds, but we function at the level of seconds and sometimes minutes. And that complicates its study, but it also makes it more beautiful. These extremes require levels of computational explanation and levels of observation at all these scales.

What conclusions could you draw from this conversation with Mara Dierssen?

To reaffirm the prevailing philosophy of today. On both sides, on the experimental side represented by Mara and on the computational side represented by me, to really apply these basic technological principles to the field of neuroscience. You can’t advance neuroscience merely with observations, you can’t advance neuroscience with computational models, you need the active synchronisation of both elements, it’s fundamental. I believe that this has been one of the great revolutions of the 21st century, originated above all by genetics and neuroscience, where being confronted with complex systems and with so much data, at some point we realised that a theoretical framework is necessary.

Gustavo, what do you love most about the brain?

Perhaps its robustness, it is an extremely robust system. Terrible things can happen, such as a cardiovascular accident or a tumour, which are irreparable damages, and yet the brain tries to compensate for them. If it had been designed by an engineer, the damage would probably be more catastrophic. Imagine a computer if you take out a piece of a chip, it stops working, right? Well, the brain continues to work, and for me that’s a great thing, that robustness, which we still don’t really understand why it’s there, but it gives the brain a character that is unique and enviable from an engineering point of view. It’s one of the issues that surprises me the most and that I understand the least, that integrity of the brain against external attacks. The integrity of the brain in the face of external attack.

And looking to the future, what discovery would be a revolution in knowledge?

A theory. Interestingly, there is still no theory of the brain. In fact, I originally came from the study of physics. I did my first PhD in quantum physics. Despite my fascination with this science and its complexity, I always had the feeling that it was relatively boring. The great years of quantum physics were between 1923 and 1925. I felt sorry that I was not around at that time when the great physicists discovered everything. They discovered quantum, formalised it, and left us with the complex but boring problems. I think we are now in those years of discovery of neuroscience and I would love to be able to contribute to the formalisation of a theory. Approximate, of course, it’s not going to be exact. But it would be amazing to be surrounded by the Newton or the Kepler of neuroscience.

Do you think these kinds of geniuses had a different brain from the rest?

I don’t know. I can answer informally, not as a neuroscientist, because there is no evidence. But informally, yes, possibly their brain has a way of integrating or segregating information that is very particular and that other people don’t have. But there is no evidence for that. It is difficult to analyse because there are not so many geniuses either, so it is very difficult to discover.

My brain is squeezed after this conversation. Intense but fascinating!

The brain is exciting and the best thing is that we are in an exciting time. We know the ingredients to understand its complexity, but we haven’t found the solution yet. I believe that when we do, it will have a huge societal impact. Not only in the clinical field, I think in many aspects. In justice, in law, in education. Understanding the brain will have an impact on all areas of society, even life. And that is also extremely exciting.