INFORMATION TECHNOLOGY

Science: “Language communication” that simulates the brain Scientists have developed new “memristors” and simulate neurochemical signals and electrical signal transduction


Humans know the world through the brain, but they know very little about the brain that knows the world.

One of the reasons is that the brain has two “languages” (electrical signals and chemical signals), and people can currently “read” the brain’s “electrical language” (read and interpret electrical signals), but they are helpless to “translate” its “chemical language” (chemical information such as neurotransmitters released by neurons).

On January 13, Yu Ping, a researcher at the Institute of Chemistry, Chinese Academy of Sciences, and Mao Lanqun’s team, together with collaborators, published a paper in Science to report a fluid memristor with polyelectrolyte confinement, and used a single device to simulate neurochemical signals and electrical signal transduction for the first time. This means that in the future, people are expected to read the “chemical language” of the brain, better simulate the brain, and realize intelligent interaction with the brain.

6380913957524202697792387.png

Schematic diagram of neuron (A) and nanofluid memristor (B) Photo courtesy of interviewee

Explore the brain “race for a thousand sails”

From artificial intelligence to brain-like intelligence, from recording signals to brain-computer fusion, from signal recognition to intelligent perception, human beings are exploring the brain and imitating its operating mechanism and function more and more deeply.

At present, many countries and regions are actively deploying research in the field of brain-like research. The European Union has launched the “Human Brain Program”, the United States has the “Brain Research Program for Promoting Innovative Neurotechnology”, and China has also opened scientific and technological innovation 2030 – “major projects in brain science and brain-like research”. In addition, Google, Microsoft and other companies have also invested a lot of money and manpower in research and development in this field.

In the field of brain-like research, the current main research focuses on the following aspects: first, brain-like intelligence, that is, neuromorphic computing, which aims to simulate the morphological structure and information processing mechanism of the human brain; The second is brain-computer fusion, which aims to simulate brain function and achieve fusion communication with the brain; The third is intelligent biomedical applications, such as neural prosthetics, intelligent sensors and intelligent sensory systems.

In these fields, scientists have done a lot of excellent work, and a large number of devices and models that simulate the structure and mechanism of brain nerves have been reported. For example, inorganic solid-state neuromorphological devices developed using two-port memristors and three-port neuroplastic transistors have achieved a series of complex computing tasks, including ultra-low-power parallel computing and establishing artificial neural networks.

At the same time, the results in the field of organic electronics research also show that neuromorphic devices based on organic materials have many potential values, especially in combination with biological systems.

Magical “Memristors”

“The nerve function of the brain is closely related to chemical and electrical signals.” Yu Ping told China Science News, “The brain generally releases neurotransmitters, including electrical signals and chemical signals, when it performs neurotransmission. At present, people can only realize the recognition and simulation of electrical signals, and it is difficult to directly perceive chemical signals. Therefore, the preparation of artificial synapses with chemical signal response (to achieve chemical-like synaptic functions) has become a scientific problem in the field of neural intelligent sensing and simulation. ”

In the 70s of the last century, scientists foreshadowed the possibility of developing a fourth element – a memristor through theoretical derivation.

“We can think of memristors as basic electronic components like capacitors, inductors, and resistors.” Yu Ping explained that as a new class of electronic components, fluid memristors are expected to simulate the brain’s “ion channel” function and realize intelligent interaction with the brain, which is expected to help people interpret the brain and realize brain-like intelligence research, brain-like computing and brain-like intelligent sensing.

Recently, scientists have used the electron current generated by the electrochemical oxidation process of dopamine to regulate neurotransistors and conductive bridge memristors to achieve the chemical regulation of synaptic plasticity function. However, chemically regulated neuromorphic devices still face many problems, for example, almost all neuromorphic devices are solid-state devices, and it is difficult to achieve chemical interaction with external signals. Therefore, the simulation of chemical signal transduction between chemical and electrical signals of synapses-like has not yet been implemented on a single device.

A critical step

In order to solve the above problems, Yu Ping and Mao Lanqun team, in collaboration with researchers from the University of Chinese Academy of Sciences, Xiangtan University and Beijing Normal University, used their long-term accumulation in the fields of brain neuroelectroanalytical chemistry and confined ion transport to propose the development of neurosynaptic function based on confined fluid devices.

On the basis of constructing the polyelectrolyte confined fluid system, the researchers found that the system has the characteristics of a memristor and uses the ions in solution to transport in the confined space of the polyelectrolyte brush, so that the device has a memory effect and successfully simulates a variety of neural electrical impulse behavior.

“Compared to traditional solid-state devices, this fluid device has a comparable operating voltage and low power consumption to biological systems.” Xiong Tianyi, the first author of the paper and a doctoral student, said, “More importantly, based on the characteristics of the fluid system, this device can simulate the regulation of memory function by neurotransmitters in physiological solution, and successfully simulate the chemical regulation behavior of synaptic plasticity.” ”

Then, the research team used the polyelectrolyte’s ability to recognize different pairs of ions to realize the simulation of transduction between neurochemical signals and electrical signals, which took a key step in the field of simulation of chemical synapses.

“The capture, regulation and simulation of chemical signals is an important part of brain chemistry research, which is very challenging and scientifically valuable. Simulating the transduction of chemical signals and electrical signals with a single device means that in the future we can ‘read’ the chemical language of the brain. If this path goes through, it will be crucial for people to understand the brain, simulate the brain, or interact with brain intelligence. Mao Lanqun said.

The team believes that this cross-field research is very exciting, which can guide scholars in the fields of intelligent sensing and neuroprosthetic perception to conduct deeper investigations. In the future, people will even be able to “save something” and “write” knowledge or skills to the brain, as in science fiction.

“At the moment we have just made a simple DEMO (prototype, prototype), which is still a concept, and it is far from real application.” “But this is an initial and crucial step for humans to intelligently interact with brain chemical signals.” (Source: Zhang Shuanghu, China Science News)

Related paper information:https://doi.org/10.1126/science.adc9150



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button