Duru, Jens

Neuronal Networks on CMOS based High Density Microelectrode Arrays (HD-MEAs)

HD-MEAs allow almost single cell resolution of electrical activity of neuronal networks. We are using MEAs from Maxwell Biosystems (❶: external page mxwbio.com, Image Source: Müller et. al - High-resolution CMOS MEA platform to study neurons at subcellular, cellular, and network levels). Shown here is an exemplary random network on a CMOS HD-MEA (❷), the spike average heatmap of a small part of the network on 9x9 electrodes (❸) and a signal recorded from a single electrode after electrical stimulation (❹). In general, we have 26’400 electrodes to stimulate on and record from.
 

Engineered neuronal networks on CMOS MEAs

My goal is to extend our successful approaches of bottom-up neuroscience to CMOS based HD-MEAs. We use small PDMS masks that control the growth of neurons to form small engineered neuronal networks. Shown here are hippocampus neurons growing within a PDMS mask.
 

Novel neuron patterning methods

I am experimenting with novel methods to pattern neurons on HD-MEAs and other surfaces. Shown here is an example of neurons that were cultured on a circle that promotes cell adhesion. While axons escaped radially, the cell bodies remained mainly on the circle.
 

Neural computation: Can we use neuronal networks to mine Bitcoins?

While Bitcoin mining probably remains a dream, exploiting neuronal networks for computational purposes might not. A potential application of bottom-up neuroscience is to teach neuronal networks to perform certain computational tasks with higher efficiency than conventional electronics.
 

Jens Duru, Joël Küchler, Stephan J. Ihle, Csaba Forró, Aeneas Bernardi, Sophie Girardin, Julian Hengsteler, Stephen Wheeler, János Vörös, Tobias Ruff. (2022) Engineered biological neural networks on high density CMOS microelectrode arrays, Frontiers in Neuroscience - Neural Technology