[Paper Review] Control of the Correlation of Spontaneous Neuron Activity in Biological and Noise-activated CMOS Artificial Neural Microcircuits

Ramin Hasani, Giorgio Ferrari+6 more|arXiv (Cornell University)|Feb 24, 2017

Neuroscience and Neural Engineering46 references3 citations

TL;DR

This paper presents a low-power, CMOS-based artificial neural microcircuit that replicates spontaneous firing correlations observed in biological cortical microcircuits by using tailored noise injection and configurable synaptic connectivity. By emulating interneuron-mediated communication with multiple synapses, the artificial system achieves correlation levels matching biological networks, revealing that high synaptic connectivity is essential for strong inter-population correlation.

ABSTRACT

There are several indications that brain is organized not on a basis of individual unreliable neurons, but on a micro-circuital scale providing Lego blocks employed to create complex architectures. At such an intermediate scale, the firing activity in the microcircuits is governed by collective effects emerging by the background noise soliciting spontaneous firing, the degree of mutual connections between the neurons, and the topology of the connections. We compare spontaneous firing activity of small populations of neurons adhering to an engineered scaffold with simulations of biologically plausible CMOS artificial neuron populations whose spontaneous activity is ignited by tailored background noise. We provide a full set of flexible and low-power consuming silicon blocks including neurons, excitatory and inhibitory synapses, and both white and pink noise generators for spontaneous firing activation. We achieve a comparable degree of correlation of the firing activity of the biological neurons by controlling the kind and the number of connection among the silicon neurons. The correlation between groups of neurons, organized as a ring of four distinct populations connected by the equivalent of interneurons, is triggered more effectively by adding multiple synapses to the connections than increasing the number of independent point-to-point connections. The comparison between the biological and the artificial systems suggests that a considerable number of synapses is active also in biological populations adhering to engineered scaffolds.

Motivation & Objective

• To understand how spontaneous neuronal activity correlations emerge in biological microcircuits at the population level.
• To develop a low-power, silicon-based artificial neural system that replicates biological correlation dynamics using noise-activated spiking neurons.
• To investigate how synaptic connectivity patterns and noise types influence inter-population correlation in both biological and artificial systems.
• To reverse-engineer the connectivity structure of biological microcircuits by matching their correlation profiles with artificial counterparts.
• To establish a scalable, reconfigurable neuromorphic platform for studying collective neural dynamics using CMOS technology.

Proposed method

• Design of compact, low-power CMOS silicon neurons, excitatory and inhibitory synapses, and both white and pink noise generators in 0.35 µm CMOS technology.
• Implementation of a ring topology with four isolated neural islands connected via interneurons to enable inter-population communication.
• Use of Gaussian white noise as input to trigger spontaneous spiking activity in both biological and artificial neurons.
• Employment of Pearson correlation coefficient to quantify statistical dependence between spike trains of neurons across islands.
• Systematic variation of interneuron connection types: single point-to-point vs. multiple synapses per connection, to assess impact on correlation.
• Simulation of artificial networks with 64 neurons and 1024 synapses, using MATLAB's corrcoef function to compute full correlation coefficient matrices.

Experimental results

Research questions

• RQ1 How does the number and type of synaptic connections between artificial neural islands affect the correlation of spontaneous spiking activity?
• RQ2 Can a CMOS-based artificial neural microcircuit replicate the correlation dynamics observed in biological cortical microcircuits?
• RQ3 What role does background noise (white vs. pink) play in modulating spontaneous firing and inter-population correlation?
• RQ4 How do multi-synapse connections compare to point-to-point connections in enhancing correlation between neural populations?
• RQ5 To what extent can the connectivity structure of biological microcircuits be reverse-engineered by matching artificial system output?

Key findings

• The artificial CMOS neural network achieved a correlation level between neural populations that closely matched that of biological cortical microcircuits when multiple synapses were used per interconnection.
• Adding multiple synapses to interneuron connections significantly enhanced inter-population correlation more effectively than increasing the number of independent point-to-point connections.
• The correlation coefficient matrix of the artificial system with triple synaptic connections (3×8 topology) closely resembled the correlation profile of the biological network, indicating high fidelity replication.
• Even with full connectivity to all neurons in the next island, the artificial system did not saturate in correlation, suggesting uncorrelated noise injection may act as a limiting factor.
• The results indicate that biological microcircuits likely employ a high number of functional synapses per interneuron to achieve strong, stable correlations.
• The use of both white and pink noise generators enabled flexible and biologically plausible activation of spontaneous spiking, supporting robust and low-power neuromorphic operation.