🧠 Invention of an artificial neuron capable of communicating with our biological neurons

Published by Cédric,
Article author: Cédric DEPOND
Source: Nature Communications
Other Languages: FR, DE, ES, PT

For the first time, an electrical communication has just been established between a living nerve cell and its artificial equivalent. In the laboratories of the University of Massachusetts, engineers have succeeded in creating an exchange of signals between a biological neuron and a synthetic neuron. This original approach allows electronics to use the same language as the nervous system, faithfully reproducing its natural electrical processes.

This breakthrough represents an important step for neural interfaces. Previous attempts faced a fundamental incompatibility: classical electronic systems used voltages that were too strong for biological tissues. The new technology reduces this difference by operating with energy parameters similar to those of our brain, laying the foundation for perfectly biocompatible communication.



The energy constraint of neural exchanges


Biological neurons form an extremely economical transmission network. Their activity relies on low-amplitude electrical signals, most often close to 0.1 volt. For a long time, artificial neurons failed to imitate this energy efficiency, requiring voltages up to 10 times higher and energy consumption 100 times greater.

This divergence created an insurmountable obstacle for integration with living tissues. Traditional electronic systems, due to their excessive energy intensity, overwhelmed biological cells and disrupted their normal activity. The high consumption was accompanied by a degradation of information, making any faithful exchange unachievable.

The technical solution was found using protein nanowires produced by bacteria. These microscopic structures, adapted to biological environments, transmit electrical signals at very low voltage. Their organic nature ensures their stability in the humid conditions specific to living tissues, unlike conventional electronic materials.


a) Diagram of an "integrate-and-fire" neuron showing excitatory current injection, current leakage, and membrane potential evolution up to threshold or extinction. Below, diagram of the formation of a metallic filament in a memristor.
b) Diagram of the memristor structure using protein nanowires, followed by TEM images showing a sparse then dense network (100 nm).
c) One thousand I-V scans measured on a memristor connected to a resistor.
d) Current response of the memristor to a voltage pulse whose amplitude changes from 120 mV to 10 mV at t = 0.2 s.


The potential of an integrated interface


The potential applications mainly concern the medical field. Neural prostheses and brain-machine interfaces could benefit from this energy compatibility. Direct exchange between electronic devices and nerve tissues opens up prospects for more targeted treatments of certain neurological pathologies, with better integration and reduced disruption of brain activity.

In the biomedical sensor sector, this innovation eliminates the need to amplify biological signals. Electronic devices could thus directly decode nerve impulses without an intermediate processing phase. This simplification would allow the design of more compact systems, less energy-intensive and more sensitive to fine variations in natural signals.

Neuromorphic electronics constitutes another important application area. The design of computing systems inspired by the human brain could achieve unparalleled energy performance. These bio-inspired processors would reproduce the parallelism and low consumption of biological neural networks, according to work published in Nature Communications.
Page generated in 0.283 second(s) - hosted by Contabo
About - Legal Notice - Contact
French version | German version | Spanish version | Portuguese version