Skip to content
Open access

Handwritten Digit classification with neural cultures is influenced by neural architecture, network dynamics, and decoding methods

Aug 2026 · bioRxiv · 3 citations · ⚡ 1 influential · 82 references
Biology

TL;DR

These findings establish that maximizing the computational potential of Synthetic Biological Intelligence, while avoiding false positives, requires a synergistic optimization of cellular identity, structural governance, and rigorous decoding logic.

Abstract

As silicon-based computing approaches fundamental physical limits, neurocomputing offers an energy-efficient alternative by leveraging the intrinsic non-linear dynamics of biological systems. To harness these dynamics, it is vital to understand the structure-function relationship governing how neural cultures process complex spatio-temporal information and how to appropriately decode the resulting neural electrophysiological activity. We investigated this utilizing a closed-loop electrophysiology platform, the CL1, to implement reservoir computing in human iPSC-derived neuronal networks. To systematically evaluate the variables driving neurocomputational capacity, we explored how cellular composition (cortical vs. hippocampal lineages), and the physical architecture (unstructured monolayers, 3D neural organoids, and modular networks confined by microfluidic devices) influenced electrophysiological properties and interacted with different decoding methodologies. Using a spatio-temporal version of a handwritten digit pattern recognition task (MNIST), we analyzed how these biological and analytical factors influenced classification accuracy. To ensure robust interpretation this required us to first demonstrated that reservoir computing decoding methods require strict artifact control and trial-based cross-validation to distinguish network computation from artifactual signal separability or temporal data leakage. Applying this validated frequency-domain pipeline, we suggest a clear functional hierarchy where structural modularity acts as a vital functional regularizer. Modular cortical cultures significantly outperformed unconstrained monolayers and organoids on MNIST. Furthermore, decoding frequency information from raw signals proved superior to typical time-bin decoding implementations. These findings establish that maximizing the computational potential of Synthetic Biological Intelligence, while avoiding false positives, requires a synergistic optimization of cellular identity, structural governance, and rigorous decoding logic. In doing so, this work provides a critical base establishing the criteria under which to evaluate neurocomputing implementations.

Read PDF

Similar papers

2026

Spiking Neural Networks for Efficient Classification of Peripheral Nerve Stimuli

Accurate electroneurographic (ENG) signal classification is a key function in biological and neural communication systems, where peripheral nerves act as noisy, bandwidth-limited information channels interfacing with implantable bioelectronic devices. In neural decoding and stimulation (ND&S) systems, ENG signals are t...

Silvia Mura, Arek Berç Gökdağ, A. Coviello et al. · 0 citations
Preprint Oct 2026

From communication to computation in neurons-on-a-chip: an in silico study of neurotopomorphic computing

Living neuronal networks transform inputs through recurrent cellular and population dynamics, yet it is unknown which network architecture supports which computation. Neurons-on-a-chip turn this question into a design problem because microchannels guide axonal growth and set the network architecture. We introduce IC$^3...

M. Barros · 0 citations
Open access Sep 2026

Harnessing the Intrinsic Dynamics of Biological Neural Networks for Reservoir Computing

The results establish neuronal cultures as state-dependent living reservoirs whose native dynamics support computation and are incorporated into a closed perception–action loop for Mario Kart control using a fixed readout.

Samitha S. Somathilaka, Jacob Clouse, Sasitharan Balasubramaniam · 0 citations
Open access Sep 2026

Shallow recurrent decoders for neural and behavioural dynamics

A SHallow REcurrent Decoder (SHRED) architecture for mapping the dynamics of population codes to individual neurons and other proxy measures of neural activity and behaviour and empirically demonstrate the capabilities of the method on a number of model organisms including Caenorhabditis elegans, mouse, zebrafish and h...

A. Rude, J. Kutz · 0 citations
Open access Sep 2026

Reservoir computing with neural activity inputs predicts behavior and neural dynamics in mouse decision-making.

Recorded neural activity from mice is used as additional inputs to a reservoir computer (RC), together with task-related inputs, enabling the RC to operate in a biologically grounded context and it is suggested that the HRC might generate activity patterns resembling unrecorded neural activity, which may contribute to...

Yutaro Ueoka, Hayato Maeda, Shuo Wang et al. · 0 citations
Open access Sep 2026

NeuralRNN: a unified framework for recurrent neural network methods in cognitive neuroscience

This work introduces NeuralRNN, an open-source Python framework that unifies these paradigms and model variants in a general-purpose pipeline, and provides an integrated and extensible model library spanning generic, structure-constrained, gain-modulated, and gated RNNs.

Hong-Hua Chen, Nai Ding · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.