Live and artificial neurons connected via the Internet

A group of scientists from universities in the UK, Germany, Italy, and Switzerland managed to develop a system for connecting artificial neurons with biological ones. They were connected via the Internet using a memristor, and three elements of the system were placed in different regions of Europe.

The basis of the brain is a group of neurons, the so-called neural networks. Individual neurons bind to each other by synapses. New technologies make it possible to develop analogues of neurons and connect them with artificial synapses. Of course, all this is at a relatively primitive level, but over time, scientists succeed in increasingly complex projects.

Well, a memristor in such an artificial network is needed in order to increase the efficiency of signal transmission from one neuron to another.

The main element of an artificial network is a semiconductor analogue of a neuron. This is a chip from millions of transistors. The chip generated electrical impulses, which first came to the memristor, and through it, through the microelectrode to the mouse hippocampal neuron. As it turned out, the electrical signal had an effect similar to the exciting postsynaptic potentials, from which neural impulses in the brain are formed.

This microelectrode played the role of a synapse, so it was called a synaptor.



Real synapses are plastic, and in order to achieve an analogy, scientists transmitted a signal to the memristor through two poles. The first played the role of presynaptic stimulation, since it received excitation from an artificial neuron. Well, the second was used as a postsynaptic input, returning a response from a natural neuron to the memristor.

By the way, the chain turned out to be quite complicated, but in the end it all worked, as expected.



In addition, the second part of the system was developed - it is needed to transmit a signal from a living neuron to a silicon one. Scientists have created a neuron registration system using the so-called local potential fixation method. After that, the pulses entered the second memristor, and then to the artificial neuron.

The result was a hybrid circuit that transmits a signal from a silicon cell to a real neuron.

As indicated above, the elements of the system were geographically separated. For example, silicon neurons were located in Zurich, memristors were in Southampton, and the culture of mouse neurons was in Padua. To transmit signals over the Internet, UDP was used.

In order to demonstrate the properties of synaptors, scientists modeled the long-term potentiation of the glutamatergic hippocampal synapses. In the resulting model, the first artificial neuron worked as a pacemaker. He produced electrical signals of a certain frequency. Memristors at the same time acted as a postsynaptic membrane, which adds the function of plasticity to the brain.

The developers programmed the memristors in such a way that they changed the resistance in response to the frequency of the discharge of the biological neuron. The latter was recorded by postsynaptic input. This is how hippocampal cell receptors work. Well, the second memristor worked in random charge mode. He gave out impulses randomly, and his activity was influenced by a natural neuron through a memristor.



As a result, a living cell showed activity, maintaining it even after a decrease in the frequency of irritation. The third element of the chain, in the end, showed enhanced spontaneous activity. If scientists reduced the rate of discharge of the pacemaker, a long-term depression developed, during which the activity of the entire system decreased.

Why and how can I use this system? According to scientists, it will help develop therapies for cardiac arrhythmias, hypertension, spinal cord injuries and Parkinson’s disease.



By the way, one of these days it became known about another option for the use of memristors. They were used by scientists of the Research Center “Kurchatov Institute” for storing and processing data. They will be useful in voice recognition systems and persons in transport and security complexes. In a few years, technologies can reach a level where such systems will be very small in size and have enormous information capacity.