Many scientists at the University of Twente are working on uncovering the mysteries of the brain: from understanding and curing brain diseases to replicating its incredible computational power. What are the most puzzling questions in brain motivated research? And what could we achieve once we have the answers?
Describing the brain
One could say that the anatomical structure of the human brain is the organ's least mysterious aspect, but it still holds its secrets. 'We have a fairly good understanding of the architecture, but we only have a limited understanding of how brains can so efficiently process information,' says Michel van Putten, professor of Clinical Neurophysiology at the UT and a renowned neurologist at the Medisch Spectrum Twente. 'There are essentially three basic types of brain cells: excitatory and inhibitory neurons and the glial cells. Neurons are excitable cells that continuously communicate mainly through chemical synapses using neurotransmitters or direct electrical communication. As neuroscientist Llinás once said: “Neurons are like people in a social network; they basically chat all the time.” But what is their language? How important are the electrical rhythms they generate? We don’t know.’
'Neurons are like people in a social network'
Furthermore, the brain function seems to be constantly hanging in the balance, explains Van Putten: 'Brains operate based on a very delicate balance between excitation and inhibition, at the edge of total chaos and a totalitarian regime, essentially preserving “democracy”. There needs to be a certain synchrony to allow efficient communication, but at the same time the different parts can’t do the same. A healthy brain is like a team. Team members need to collaborate and communicate, but they can’t all do the same tasks at the same time.’ Although the team comparison is accurate, there is probably no single ‘boss’ in the brain. As far as we know, brain is self-organized. There are critical regions, sometimes called ‘hubs’. These regions are extremely important and can’t be damaged without severely affecting function. On the other hand, there are areas in the brain that can be partly removed without significant consequences. Hence another puzzle without a clear solution. Brains are redundant. Why would nature create a brain with seemingly unnecessary parts?
Materials that can learn
Besides the brain’s architecture, another fundamental question worth exploring is: How can the brain accomplish so much computational power using so little energy? This question is directly addressed at the UT. Professor of Physics Hans Hilgenkamp is aiming to see if we can at least partly mimic the brain’s energy efficiency. ‘In its abilities, the brain is comparable to a supercomputer,’ says Hilgenkamp. ‘However, the brain uses million times less energy. It uses about 10-20 watts while a supercomputer needs roughly 10 megawatts. Of course, there is a difference between the two. A supercomputer can perform different tasks, it is very good in calculating, for instance. The brain, on the other hand, is much better in pattern recognition. If you look outside, you can instantly tell that there is a tree. You don’t have to calculate the amount of leaves to know. We have an amazing capacity in this regard.’
'The brain is comparable to a supercomputer'
Within the setting of the newly established BRAINS (Brain Inspired Nano-Systems) Centre, various UT scientists would like to emulate this capacity in materials. ‘There already are semiconductors that try to simulate the pattern recognition of the brain, but trying to simply mimic the function of the brain is very energy inefficient,’ says Hilgenkamp. ‘It would be much more favourable to focus on the hardware of the brain. Hardware is fixed in a computer, but in the brain it’s always adapting. The brain’s hardware uses connections that develop every time you learn something new. This is what they call the plasticity of the brain. We therefore want to create materials that can develop themselves as well, that can learn. We could call them learning materials, also known as neuromorphic computing.’
What is holding the researchers back from realizing such materials? ‘There are many things in the brain that we understand but that are too complex to mimic in electronic materials,’ answers Hilgenkamp. ‘In a computer, a logic part and memory are separated in space. This is what we call the Von Neumann bottleneck. You constantly have to transfer the information from one to the other. In the brain, these two parts are intertwined, there is a colocation of memory and logic - which is something we can’t recreate yet. Another complication is that the brain is very three-dimensional. Each neuron is connected with thousands of others through a huge 3D network. It’s difficult to realize the same structure with the current technology. Semiconductors are based on thin layers. If we can make the step towards 3D structures, that would be a huge accomplishment.’ It’s precisely this 3D structure that Hans Hilgenkamp believes to be the reason why the brain is able to achieve its exceptional computational power - and the reason why current technology can’t compete with it.
If we could recreate the power of our own brain, it could revolutionize the world of electronics. For example, we could apply learning materials in cameras for autonomous driving where pattern recognition is imperative. Regular computers and other electronics could become more energy efficient. ‘A significant fraction of all energy worldwide is used for information technology. We could reduce that,’ Hilgenkamp points out. Learning materials would naturally also be useful for AI (artificial intelligence) development. ‘AI can be implemented in a normal hardware in a normal computer, but that costs an enormous amount of energy.’ This would be solved - if we knew how the brain works.
Fixing the brain
Once the mysteries of the brain become unravelled, we can do more than improve the world’s energy consumption. We can ‘fix’ ourselves. Because our brain seems as vulnerable as it is powerful. Brain diseases are something we all hope to avoid, yet they occur in high numbers. And more often than not, their cause and mechanisms remain an enigma.
‘Besides acute disorders, such as stroke or traumatic brain injury, you could divide brain diseases into two groups,’ Michel van Putten explains. ‘Several neurological disorders are characterizes by cycles, such as epilepsy or migraine. In these cases, certain transitions can occur in otherwise normally functioning brains. For instance, in many patients with epilepsy brain function is nearly always normal. However, “glitches” can occur, resulting in a seizure – which suddenly begins and suddenly ends. Patients with epilepsy alternate between these two conditions and most of the time we have no clue what the cause is. The second category are slowly progressing diseases such as Parkinson's or dementia, which are caused by progressive failure of the neural networks. Such progressive neurodegenerative diseases are often irreversible, in part because neurons will eventually die if deprived from sufficient input, and unlike other bodily tissues, neurons cannot regenerate.’
What triggers these ‘mishaps’ in the brain is largely unknown and that means they are not preventable. Professor of Nanobiophysics Mireille Claessens is contributing to changing that. ‘I focus on molecular understanding of brain processes. It’s the mechanisms that destroy cells that are at the core of diseases,’ says the expert. ‘Protein aggregations are at the base of many neurodegenerative diseases, such as Parkinson’s, Alzheimer, dementia, ALS. We know that in case of Parkinson’s and Alzheimer’s, protein plaques are the hallmarks of the disease. What we don’t know is why they form.’
Claessens’ current research concentrates on understanding the mechanism behind Parkinson’s disease. ‘The normal role of the protein associated with Parkinson’s is not clearly defined,’ she says. ‘It has no clear 3D structure and probably many different functions. We don’t know what exactly this protein does. Moreover, it’s not an essential protein. You can completely knock it out of the organism and the organism will survive. The mystery is understanding how this protein works. Because there must be advantages to having it. Considering that it poses a risk, you would expect that evolution would have got rid of it. However, aggregation of the protein happens long after the reproductive age. It is a problem because the society is getting older. Age is the main risk factor when it comes to Parkinson’s. We are basically fighting something that is not an evolutionary problem. We are trying to solve something that nature didn’t deem necessary to solve. And we don’t even know if it’s solvable.’
'We are trying to solve something that nature didn’t deem necessary to solve'
Should we do everything we can?
Regardless of these doubts, Claessens notes that curing brain diseases would be wonderful for all affected individuals. Philosopher Saskia Nagel agrees, but suggests to be cautious when moving forward: ‘It could be fantastic for individuals and their families. At the same time, we need to be sensitive on the way to get there. We need to have a shared understanding of what we think we should cure before we go on intervening. This is a critical task for responsible research. Should we do what we can do? With progress in the neurosciences, this is particularly relevant as it touches the core of what we think makes us human. While treating diseases is promising in most cases, there are also grey areas where we cannot say for sure that something should be treated as a disease. Note that our concept of what counts as diseases changes – homosexuality was once understood as a disorder, after having been understood as a sin before. It is wise to be careful with what we aim to cure.’
Mathematics to the rescue
While we are far away from curing or preventing Parkinson’s, we are able to relieve its symptoms. Besides medication, doctors are using deep brain stimulation. This method involves opening the patient’s skull and installing an electrode deep inside so that the brain is stimulated continuously. Professor Stephan van Gils from the UT department of Applied Mathematics believes that this method isn’t as efficient as it could be. ‘In many cases this works miracles, but in some cases the response is sub-optimal or even accompanied by more problems. We want to make a mathematical model to test other types of stimulation. We will test the deep brain stimulation devices on artificial basil ganglia developed in Nijmegen. It is quite spectacular,’ he says enthusiastically.
Van Gils is a mathematician aiming to uncover some of the mysteries of the brain. Specifically, he’d like to further the understanding of pathologies of the brain - of why, how and where they originate. He is convinced that mathematics is the one way to find these answers. ‘I believe mathematics can contribute to crucial understanding of the brain. In fact, mathematics is the only science that can help us understand how this black box works. It can model the complexity. It can extract the abstract rules that rule the dynamics. I believe that one day mathematics can even help us understand the cognitive functions of the brain.’
Finding the epilepsy signature
For now, the mathematician mainly works on developing dynamic models of the brain to help make decisions in medical practice. One of his research projects focuses on epilepsy. ‘More specifically focal epilepsy,’ he clarifies. ‘If patients with this epilepsy don’t respond to medication, surgery is required. Our aim is to figure out what part of the brain can be taken out, so the patient becomes seizures free. Our partner University Medical Center Utrecht is trying to figure out how different parts of the brain are connected. Our task is to translate the outcome of their measurements to models for brain activity. Based on those models we can then predict which part of the brain can be removed in order to function optimally. We are making progress and it looks very promising.’
Michel van Putten is also working towards better diagnostics and treatment of epilepsy. ‘We are trying to improve the interpretation of the electroencephalogram (EEG), that may show the specific ignition point of epilepsy. EEG is a monitoring method to record electrical activity of the brain. In essence, it allows you to listen to the neurons’ talking, reflected in a rich repertoire of brain rhythms. In case of epilepsy, you are able to see that certain neurons don’t chat correctly. For instance, you can see a little spike when the neurons basically talk too loud. Sometimes they only talk loud but don’t disturb the neighbours; if they do, however, it may lead to a seizure. Interesting questions include: is it the neighbours that suddenly listen better, or is it the talking that is too loud, or both? Further, it takes a long time to assess the recordings visually, which is why we are using deep learning to detect anomalies in the EEG. We aim to find the epilepsy signature and hopefully contribute to a much faster diagnosis.’
You’ve caught me in a coma
Van Putten has more examples of how EEG can lead to ‘saving our brains’. Using the EEG recordings of patients who suffered cardiac arrest, scientists can study energy deprived brains and possibly provide a better treatment for comatose patients.
'We may be able to shift the line of irreversible damage'
‘Brains need energy and are absolutely dependent on blood flow. After the heart stops, you can see then it only takes about twenty seconds for the brain to stop working as well. While this is still completely reversible for about two to three minutes, if it lasts longer synapses and neurons may become irreversibly damaged. However, the point of no return is not clear. This is what we study on a more fundamental level in our in-vitro lab at the UT, using a ‘brain-on-a-chip’ model to simulate this clinical condition. Important questions include: What defines the tipping point? Why do some neurons recover and others not? Can we limit damage by therapeutic interventions? There is so much to learn, and it is our hope that we may be able to shift the line of irreversible damage.’
Analyzing the brain
A fair amount of brain related mysteries has already been mentioned, but when asked about which mystery is the biggest of all, the scientists all answer similarly: our consciousness, our self-awareness, feelings, memory, language... In other words, the cognitive functions of the brain - possibly the part that makes us truly human. ‘It’s just fascinating that nobody has a clue of what happens in our brain when we say one plus one is two. Nobody has a clue how we think,’ says Stephan van Gils.
It is true - nobody knows how we think. But we can see when we think. Once again, thanks to EEG. ‘In my work I mostly use EEG to see how it can help us understand the relation between cognitive functions and the brain,’ says UT-based cognitive psychologist Rob van der Lubbe. Although measured brain activity is generally very difficult to interpret, there are certain things we are able to ‘read’ from EEG, says the Associate Professor. ‘We can observe changes in brain activity when people are thinking and when they are performing motor functions. There is even something we call motor imagery: the motor areas in the brain show a change in specific frequency bands even if you are just thinking of the motion. The brain activity is actually almost the same when the person is only imagining the motion as when she/he is carrying out the task. In the long run, these findings could be beneficial for treatment of patients after stroke, for example, as people could train motion even without doing anything.'
Can we measure the moment when people decide to do something? ‘Is there even such a moment?’ asks Van der Lubbe in return. ‘Scientists like Benjamin Libet performed experiments in the past, asking people to make a hand movement at any moment and then tried to determine when exactly was the decision made. The EEG findings suggested that the participants decided to make the movement about 500 milliseconds before they said they did. But is this really what the measurements show? What they measured could have been just the preparation of the brain to do something. Is there really a moment when we make a conscious decision? That is hard to answer because we don’t even know what consciousness is.’
'Consciousness is like “fame in the brain”.'
‘The relation between consciousness and brain activity is not clear,’ he continues. ‘The brain is in a constant interaction with its environment, it’s never a passive receiver. What is there around us and how we perceive it is not the same. Incoming information is already modulated by the brain based on existing information within the brain, based on our experiences, memory and knowledge. That means that probably everyone perceives the world rather differently. The relationship between brain activity and consciousness is definitely the big mystery I would like to solve. There are several theories but I don’t think any of them are true. I think it’s too simplistic to localize consciousness in a specific brain area. I think it’s more likely that your level of consciousness depends on whether information can go to all areas of the brain. Maybe consciousness is as Daniel Dennett once said, like “fame in the brain”.'
Are we too quick to dismiss the option that consciousness is not even fully brain dependant? Saskia Nagel suggests exploring further. ‘If it is all just firing neurons, how can there be such differences between experiences? Why is one experience, for example seeing the trees outside, so different from another experience like feeling the sand at the beach with my feet?’
Using the brain
So many secrets remain concealed inside our skulls. Yet there are so many clever brains working on unveiling them(selves), it is not unrealistic to imagine a world in which brain diseases have been eradicated, brain inspired AI has access to all and brains are regularly modified to make us think or feel differently. What would that mean for us?
‘In such a future, population would be able to grow much older, but we could argue whether that is necessarily a positive thing,’ says Mireille Claessens. ‘Just think of the innovative power that may be hampered. After a certain age your ideas are mainly based on experience, not on bright insights. If you look at Nobel Prize winners, most of them won thanks to work they did when they were relatively young – that is when they got their crazy idea.’
'There are always some threats'
Hans Hilgenkamp doesn’t have a straightforward answer. ‘Imagine someone asked you this forty years ago. Imagine they said that we would have super powerful computers, brain scans, facial recognition… That would sound quite scary. And yes, of course there are threats coming hand in hand with new technologies. Cybercrime wouldn’t exist without computers. But there are always some threats. Forty years ago there were no hackers but there was the Cold War. The question is: does the technology only create these threats or can it also help to mitigate them? For example, you can use AI to detect cybercrime in very early stages. Still, we should think of these possibilities as researchers. We shouldn’t close our eyes to how the technology that we are developing could be used.’
Saskia Nagel offers a similar thought. She believes that before revealing the mysteries hidden inside our heads, we should use these heads to consider whether it will truly help us to know the answers. ‘It would first and foremost require that we deeply think about what we want it to mean for us,’ says the philosopher. ‘We need a discussion about which values we want to protect. What do we value and why, and which boundaries do we want to keep? Which knowledge will help us reach our goals? This is a question we should ask in the process of scientific discovery. It will be central for our individual and social well-being to have a good idea of how we use the brain related knowledge that we gain. Just think of this line from time to time: “With great power comes great responsibility“.’
experts who contributed to the article:
- Michel van Putten, Professor of Clinical Neurophysiology, TNW Faculty
- Mireille Claessens, Professor of Nanobiophysics, TNW Faculty
- Stephan van Gils, Professor of Nonlinear Analysis, EEMCS Faculty
- Rob van der Lubbe, Associate Professor of Cognitive Psychology, BMS Faculty
- Hans Hilgenkamp, Professor of Applied Physics, TNW Faculty
- Saskia Nagel, Associate Professor of Philosophy and Ethics, BMS Faculty and Professor of Applied Ethics RWTH Aachen Univers
You can also find this story in our brand-new Science & Technology Magazine. Grab a printed copy on campus or see the whole magazine here.