Max Plank researchers have developed bio-realistic artificial neurons that can work in a biological environment and produce diverse spiking dynamics

The development of neuromorphic electronic devices depends on an accurate mimicry of neurons. Artificial neurons can’t operate in biological environments. The creation of organic artificial neurons based on circuit oscillators has been made. These require a large number of elements to implement. A compact nonlinear element based organic artificial neuron has been described. This artificial neuron can operate in liquids and is sensitive to its surrounding biological species. The system provides in-situ operations, spiking behaviors, and ion specifity under biologically relevant conditions including normal physiological concentrations and pathological ones. The neuronal excitability is regulated by variations in biomolecular and ionic concentrations, while small-amplitude oscillations in the electrolytic medium and noise alter its dynamics. In a biohybrid, an artificial neuron works in synergy with epithelial and membrane cells.

The basic units of nervous system, neurons are used for transmitting and processing electrochemical signals. They communicate through gaps between the axons of presynaptic and postsynaptic neuron dendrites in a liquid-electrolytic medium. Neuromorphic computing is a hardware-based solution that mimics the behavior of neurons and synapses. Microelectronics can mimic neuronal behavior by using oscillatory topologies. These approaches, however, can only mimic specific aspects of neuronal behaviour by integrating passive electronic components and many transistors, resulting a bulky biomemtic system that is unsuitable for in situ direct biointerfacing. Nonlinear and volatile devices, such as spin torque oscillators and memristors, can be used to increase integration density and mimic neuronal dynamics.

Source:

Max Plank AI Researchers Have Developed Bio-Realistic Artificial Neurons That Can Work In A Biological Environment And Can Produce Diverse Spiking Dynamics