Modeling the brain as an electrochemical machine
Electroencephalography (EEG) can be compared to hanging microphones high above a football stadium and analysing the roars from spectators, in order to detect what’s going on down on the field. Imagine if we instead could interview the individual spectators, meaning the neurons in the brain?
The thought may seem strange to a lot of people, but it is actually a fact that everything you think, feel, sense, perform or dream about is communicated between neurons in your brain as a combination of electrical and chemical signals. In short: The brain is an electrochemical machine.
The first measurement of the electric activity in the human brain was performed by the pioneering German psychiatrist Hans Berger more than 90 years ago, but he didn’t really know what he was measuring.
Because of this uncertainty, Berger waited five years before he published the first scientific paper about his method for measuring the electrical activity inside the brain with the help of electrodes on the outside. The method was initially met with disbelief and ridicule, and it took a long time before electroencephalography (EEG) was developed into the essential tool it is today.
EEG can be developed further
Today, EEG is used in hospitals to survey the brain of patients with as diverse illnesses as epilepsy, sleep disorders, eating disorders, coma, and so on. But the professors Gaute Einevoll and Anders Malthe-Sørenssen at the Department of Physics at the University of Oslo (UiO) are convinced that the technology can be developed a lot further.
The Research Council of Norway has recently granted 9 MNOK to a project led by Einevoll, Malthe-Sørenssen and associate professor Marianne Fyhn at the Department of Biosciences at UiO. The project aims to investigate the relationship between the electrical activity inside the brain and the signals that can be registered on the outside.
The project is called COBRA – COmputing BRAin Signals – and is groundbreaking in the way that physicists and mathematicians are heavily involved in brain studies. Most neuroscientists, both in Norway and internationally, until now had their background in biology, medicine or psychology.
“We physicists have a lot of knowledge about electromagnetism and general modelling, and we are really looking forward to cooperate with the biologists who know the physiology of the brain much better than we do ourselves. I would like to emphasize that we are very happy at the Department of Physics, because we were able to recruit Gaute Einevoll. He has extensive knowledge about both physics and neuroscience”, states Anders Malthe-Sørenssen.
He is referring to the fact that Einevoll was hired as Adjunct Professor at the institute in 2014. Einevoll is an expert in Computational Neuroscience, an area which focuses on mathematical modelling of the brain and the nervous system. Malthe-Sørenssen himself is an expert in Computational Physics and has been a key figure in developing both science and teaching on this topic at the UiO.
Roaring tells what happens in the stadium
“State of the art EEG-mapping today can be compared to hanging microphones high above a football stadium, and then analysing the roaring and the moaning in the audience in order to understand what the football players are doing on the court. For instance, it is reasonable to assume that the roars are louder when the home team scores a goal. But we are convinced that we can do much better than this by using the technique called forward modelling”, explains Einevoll.
The forward modelling in project COBRA starts with the researchers building a digital model of a part of the brain.
Then, they are going to simulate a specific activity in a network of neurons in the model. The next step is to calculate the EEG signals this activity will give rise to at the outside of the skull. The next step after this is to compare the predictions of the model with measurements from a real brain, and this can be used to adjust the digital model until it fits nicely with reality.
“We are for instance going to do experiments with mice when they are repeating the same task over and over again, while we shall be mapping the neuronal networks that are active in their brains and registering the EEG signals that arise”, explains Einevoll.
“To continue using the metaphoric football stadium, this is like asking Norway’s football player, Ole Gunnar Solskjær, to score the same goal over and over again. Until we are certain that we are able to recognize the exact roar of excitement that arises when he scores for the home team”, explains Malthe-Sørenssen.
The brain is an electrochemical machine
The rationale behind the EEG technology is that the human brain contains ca. 100 million neurons, each of them in principle a small electrochemical machine. By extension, the whole brain is also an electrochemical machine.
The neurons are connected in enormous networks. An “average” neuron can communicate with approximately 10 000 other neurons.
A neuron that is going to send information to another neuron, starts by sending an electrical signal through a thin nervous wire – an axon – which is actually a part of the neuron. This electrical signal cannot jump directly to the next neuron, but instead causes a small chemical signal to be released in the point of contact – the synapse – between the neurons.
The receiving neuron registers the chemical signal and converts it to a new electrical signal, which is computed in the neuron’s central body. This computation sometimes gives rise to a new electrical signal, which is passed to yet another neuron.
Electromagnetism as described by the Scottish physicist and mathematician James Clerk Maxwell in 1865, tells us that every electric current gives rise to both electric and magnetic fields in the vicinity.
This happens also in the brain: The electric current in the neurons gives rise to electrical and magnetic fields that can be registered on the outside of the brain with the help of either electroencephalography (EEG) or magnetoencephalography (MEG).
The electrical signals transmitted between neurons are called action potentials. They have a duration of about one thousandth of a second, and the voltage difference across the cell membrane is approximately 0,1 volts. The signals have typically only one thousandth of this strength at the outside of the cell, but it is nevertheless no big problem to register them a few centimetres away.
The problem is that the brain is able to fire signals from millions of neurons at the same time, and this makes it very hard to establish exactly which neurons – or networks of neurons – that are “yelling”.
Impossible to interview individual neurons
“You can imagine us lowering the microphones from above the stadium and down to the stands in order to interview individual spectators, that is our metaphoric neurons".
"But it is not feasible to insert electrodes into peoples’ brains, except for in very special cases when necessary for medical treatment. That’s why we use mice brains and modelling instead”, explains Einevoll.
The scientists in project COBRA cooperate with scientists in another project called DigiBrain, led by associate professor Marianne Fyhn. The vision of DigiBrain is to generate more knowledge about mental illnesses and, in the long term, contribute to the development of better drugs.
“One of the aims of DigiBrain is to search for differences in neuronal networks in the brains of healthy people and schizophrenics, and make comparisons. We are already able to spot some differences using EEG, but we need project COBRA in order to understand the link between the electrical activity inside the brain and the signals we are able to measure on the outside”, explains Einevoll.
Both projects, COBRA and DigiBrain, are affiliated the UiO’s Centre for Integrative Neuroplasticity (CINPLA), where scientists study neural plasticity at multiple levels.
Mind reading – is it possible?
“Will neuroscientists soon be able to read peoples’ minds?
“No, that’s not something you should be worried about! Admittedly, we have taken a few steps in that direction. For instance, it is possible to register the signals in motor neurons in the brains of people with paralyzed limbs, and transmit these signals to a robotic limb”, Einevoll answers.
“We should also be able to spot the different patterns of activity in your brain when you are, for instance, playing tennis or walking around in your apartment. But we would have to attach electrodes to your skull in order to perform such tricks, and this would obviously draw your attention. There is at present no technology that allows us to “read the mind” from a distance, and I can not imagine that such a thing would ever be possible”, adds Einevoll.
Mest lest siste syv dager
Nyheter fra andre
Kunnskap om hjernen kan gjøre kunstig intelligens bedre. Samtidig kan kunstig intelligens brukes til å lære mer om hjernen vår.
With the powerful eyes of the Hubble Space Telescope, an international team of astronomers led by Håkon Dahle at UiO observed straight into the bright and hot heart of a galaxy 11 billion years old in no less than 12 multiple, gravitationally lensed images.
Ingen kommer på innsiden av kunstig intelligens. Kunstig intelligens er dessuten lett å lure. Ny matematikk kan ta innersvingen på begge utfordringene.
De har utviklet en solcelle som bruker bakterier for å produsere strøm. Og ble belønnet med sølvmedalje i internasjonal bioteknologikonkurranse.