Brain-Computer Interfaces, what can and cannot do?

In this article we will talk about BCI, Brain Computer Interface, also called neural interfaces, that is, those devices that allow us to capture the messages of our brain and transfer them to a computer, which in turn can control other devices. Neural interfaces have been in development for decades and you may have heard of Neuralink, Elon Musk's company that recently made headlines because a prototype of it allowed a monkey to play Pong. How do neural interfaces work and what do they really do? Will they allow us to download data into our brains as we saw happen in Johnny Mnemonic or The Matrix, so that we can learn to drive a helicopter in a matter of minutes? How does a neural interface work? A neural interface essentially reads the electrochemical activations that take place between the synapses of the brain, interprets them and translates them into a digital signal that can be processed by a computer. When you make a movement, such as raising your arm, something happens in the brain in the form of an electrochemical signal. These signals, in practice electron movements, can be detected by devices connected to the brain and transformed into data that are sent to a computer, where a software will take care of identifying the type of signal, translating it into some action. For example, when you raise an arm in your brain, a certain activation occurs. The software recognizes its specificity and matches it with the movement of a mechanical arm connected to the computer itself. This is, explained in the simplest possible way, what a BCI can do today.

What types of neural interfaces are there? There are two types of interfaces, which are categorized according to the way they collect the electrical information emitted by the brain: the external ones, such as the classic headphones where there are multiple cables connected to a computer, and the internal ones, that is, implanted directly on the surface of the brain. The main difference is that the external ones are clearly less invasive, but, not being in direct contact with the brain, the amount and accuracy of the information they can locate is low. To give you an example: imagine you are outside a room where there are a hundred people talking. What you will hear will be a mix of many sounds; maybe you will be able to identify some characteristics, such as the presence of female voices or laughter, but nothing specific and detailed. If, on the other hand, you enter that room, the amount of information you will be able to identify will be much higher: you will be able to hear the individual voices and better understand what they say. Obviously entering the room, that is, entering the brain, allows you to have a lot more information to use. These interfaces are currently the domain of the medical field, since most of the studies have been done to allow paralyzed patients to achieve greater freedom of movement. One example above all: an external system based on the electroencephalogram (EEG).

External systems based on the electroencephalogram allowed paralyzed patients to return to walking again thanks to robotic prostheses. Earlier we told you that when you raise an arm, something happens in your brain. In reality the opposite happens, that is, unconsciously one thinks of making a movement, so to that thought in the brain an electrochemical activation takes place, to which the movement of the arm corresponds. To create that electrochemical activation it is not necessary that an arm is really present, since the movement of the arm is a direct consequence. This means that, even if you did not have an arm, the thought of making that movement would still create an electrochemical reaction, which in this case would be detected by the neural interface, which, in turn, would activate the robotic arm to perform that action after the signal has been processed by the computer. Clearly it is not so easy, patients who are in this condition have to train the brain, but it has been shown that it is possible to "learn to walk again" by doing this type of mental training. The most invasive systems connected directly to the brain allow to detect many more signals than external ones. Having many more signals available means being able to identify, interpret and manage more complex movements. The information needed to move the arm upwards is certainly less than a more complex movement, such as that of all fingers, for which an internal interface would certainly be more effective. What is the future of neural interfaces? At the beginning of 2021, Brown University showed a prototype of internal wireless BCI that relied on a wireless broadband link, representing a great technological advance since it can be implanted in patients without them having to be physically connected to a computer. There is, however, a disadvantage for systems connected directly to the brain and even more disadvantages for wireless ones. In fact, these solutions must be able to interpret a lot of data and consequently be complex, needing more energy to work. As a result, the challenge now shifts to miniaturization and efficiency of devices.

What has been feared by Elon Musk or science fiction films, with neural interfaces capable of allowing access to memories or loading capacity inside the brain, seem to belong very much to the field of imagination and science fiction, rather than to a possible future. There are those who say that they are just pure fantasy and those who do not really close all the doors, although it is certain that, if it ever happens, it will take many years. The future of this technology is however promising with regard to the medical field, giving the possibility of having prostheses that allow veterans of accidents or debilitating diseases to regain mobility. Paralysis of a limb does not correspond to damage to the brain and the latter can be used to activate BCIs that allow, in a certain sense, to circumvent the problem. Now we need to find ways to make these interfaces, computers and robotic arts available easily, not just as part of experiments within laboratories and universities.

by Andrea Ferrario Thursday, June 16, 2022 5:00 PM