A digital implementation of neuron-astrocyte interaction for neuromorphic applications

Area of Science:

• Neuroscience
• Computational Neuroscience
• Neuromorphic Engineering

Background:

• Astrocytes are increasingly recognized for their crucial roles in neuronal activity modulation and information processing.
• Understanding neuron-astrocyte interactions is vital for advancing neuromorphic computing architectures.
• Existing models often lack efficient digital implementations for studying these complex dynamics.

Purpose of the Study:

• To propose and validate a novel digital neuromorphic circuit for simulating neuron-astrocyte interactions.
• To investigate the influence of astrocyte signaling on neuronal firing patterns using a digital model.
• To explore the potential of this circuit for enhancing information processing in neuromorphic devices.

Main Methods:

• Developed a digital circuit incorporating the Izhikevich model for neuron dynamics and a Postnov functional model for astrocyte calcium dynamics.
• Employed Single Constant Multiply (SCM) technique and linear approximations for efficient hardware implementation.
• Validated the circuit through MATLAB simulations, hardware synthesis, and Field-Programmable Gate Array (FPGA) implementation.

Main Results:

• Demonstrated that the digital astrocyte model can alter neuronal firing patterns via bidirectional communication.
• Showcased that astrocyte signaling significantly regulates information processing within synaptic clefts.
• Confirmed that the neuron-astrocyte crosstalk circuit generates diverse neural responses, boosting neuromorphic circuit capabilities.

Conclusions:

• The proposed digital neuron-astrocyte circuit effectively models biological interactions and their impact on neural computation.
• This implementation offers a low-cost, efficient platform for reconfigurable neuromorphic devices mimicking brain circuits.
• The findings highlight the potential of incorporating astrocyte dynamics to enhance the information processing capacity of artificial neural systems.