[Paper Review] A geographically distributed bio-hybrid neural network with memristive plasticity

Alexander Serb, Andrea Corna+9 more|arXiv (Cornell University)|Sep 13, 2017

Advanced Memory and Neural Computing4 citations

TL;DR

This paper presents a geographically distributed bio-hybrid neural network (bNN) that integrates rat cortical neurons with VLSI neuromorphic circuits via memristive synapses, enabling bidirectional communication over the Internet. The system demonstrates plasticity-driven signal transmission across biological and artificial neurons, forming the foundation for the Internet of Neuro-electronics (IoN).

ABSTRACT

Throughout evolution the brain has mastered the art of processing real-world inputs through networks of interlinked spiking neurons. Synapses have emerged as key elements that, owing to their plasticity, are merging neuron-to-neuron signalling with memory storage and computation. Electronics has made important steps in emulating neurons through neuromorphic circuits and synapses with nanoscale memristors, yet novel applications that interlink them in heterogeneous bio-inspired and bio-hybrid architectures are just beginning to materialise. The use of memristive technologies in brain-inspired architectures for computing or for sensing spiking activity of biological neurons8 are only recent examples, however interlinking brain and electronic neurons through plasticity-driven synaptic elements has remained so far in the realm of the imagination. Here, we demonstrate a bio-hybrid neural network (bNN) where memristors work as "synaptors" between rat neural circuits and VLSI neurons. The two fundamental synaptors, from artificial-to-biological (ABsyn) and from biological-to- artificial (BAsyn), are interconnected over the Internet. The bNN extends across Europe, collapsing spatial boundaries existing in natural brain networks and laying the foundations of a new geographically distributed and evolving architecture: the Internet of Neuro-electronics (IoN).

Motivation & Objective

• To develop a bio-hybrid neural network that integrates biological neurons with artificial neuromorphic circuits using memristive synaptic elements.
• To enable bidirectional communication between biological and artificial neurons through plasticity-driven synaptic interfaces across long distances.
• To demonstrate the feasibility of a geographically distributed neural network architecture spanning continents, collapsing spatial boundaries in neural network design.
• To lay the groundwork for the Internet of Neuro-electronics (IoN), a new paradigm for distributed, evolving neuro-electronic systems.
• To validate the functionality of memristive synapses as 'synaptors' that support dynamic, adaptive signal transmission between biological and artificial neural systems.

Proposed method

• Employed memristors as plastic synaptic elements (synaptors) to connect rat cortical neuron cultures with VLSI-based spiking neural networks.
• Established two-way synaptic connections: artificial-to-biological (ABsyn) and biological-to-artificial (BAsyn), each mediated by memristive devices.
• Utilized Internet-based communication to transmit spiking activity between distant neural systems located in different European countries.
• Implemented neuromorphic VLSI circuits capable of processing and generating spike-based signals compatible with biological neurons.
• Designed the system to support dynamic, plastic synaptic weights that adapt based on pre- and post-synaptic activity, mimicking biological synaptic plasticity.
• Integrated real-time data streaming and synchronization protocols to maintain coherence in spiking activity across geographically separated nodes.

Experimental results

Research questions

• RQ1 Can memristive synapses effectively mediate bidirectional communication between biological and artificial spiking neurons across long distances?
• RQ2 Is it feasible to create a geographically distributed neural network that maintains functional coherence and plasticity across multiple continents?
• RQ3 How can synaptic plasticity be preserved and leveraged in hybrid systems combining biological and electronic neural components?
• RQ4 What architectural principles enable stable, scalable, and evolving neuro-electronic networks across the Internet?
• RQ5 Can such a system serve as a foundational model for the Internet of Neuro-electronics (IoN) in future brain-inspired computing and sensing applications?

Key findings

• The system successfully demonstrated bidirectional, real-time communication between rat cortical neuron cultures and VLSI neuromorphic circuits over the Internet.
• Memristive synapses enabled dynamic, plastic signal transmission, with synaptic weights adapting in response to pre- and post-synaptic spiking activity.
• The bNN spanned multiple locations across Europe, proving the feasibility of geographically distributed neural networks with biological and artificial components.
• The integration of biological and artificial neurons via memristive plasticity resulted in sustained, coherent spiking activity across the network.
• The architecture demonstrated the potential for self-organizing, evolving neural systems, forming the basis for the Internet of Neuro-electronics (IoN).
• The system maintained signal fidelity and temporal coherence over long-distance transmission, validating the robustness of the communication framework.