Assessing Pattern Recognition Performance of Neuronal Cultures through Accurate Simulation

TL;DR

This paper introduces a digital simulation model of neuronal cultures to evaluate their pattern recognition capabilities, specifically in handwritten digit recognition, and suggests ways to optimize real culture performance.

Contribution

It presents a novel methodology for assessing neuronal culture performance using accurate simulations and identifies parameters that improve recognition tasks.

Findings

Simulated neuronal cultures can recognize handwritten digits.

Optimized simulation parameters enhance recognition accuracy.

The methodology guides the development of better real neuronal cultures.

Abstract

Previous work has shown that it is possible to train neuronal cultures on Multi-Electrode Arrays (MEAs), to recognize very simple patterns. However, this work was mainly focused to demonstrate that it is possible to induce plasticity in cultures, rather than performing a rigorous assessment of their pattern recognition performance. In this paper, we address this gap by developing a methodology that allows us to assess the performance of neuronal cultures on a learning task. Specifically, we propose a digital model of the real cultured neuronal networks; we identify biologically plausible simulation parameters that allow us to reliably reproduce the behavior of real cultures; we use the simulated culture to perform handwritten digit recognition and rigorously evaluate its performance; we also show that it is possible to find improved simulation parameters for the specific task, which can guide the creation of real cultures.