Engineered Biological Neural Networks on High Density CMOS Microelectrode Arrays

Area of Science:

• Neuroscience
• Bioengineering
• Materials Science

Background:

• Bottom-up neuroscience utilizes engineered neural networks for studying information processing.
• Current limitations in microelectrode arrays (MEAs) restrict network complexity and spatial resolution.

Purpose of the Study:

• To overcome challenges in integrating polydimethylsiloxane (PDMS) microstructures with high-density complementary metal-oxide-semiconductor (CMOS) MEAs.
• To enable the creation of complex, sealed, and functional in vitro neural networks with high spatial resolution.

Main Methods:

• Stamp-transferring a thin PDMS film onto CMOS MEAs to seal microstructures and prevent axon escape.
• Utilizing impedance-based methods to visualize microstructure placement and confirm axonal confinement.
• Employing high-density CMOS MEAs for sub-cellular neuronal readout and stimulation.

Main Results:

• Achieved 23% of structurally complete and sealed networks on CMOS MEAs.
• Demonstrated 45% of these networks exhibited spiking activity across all channels.
• Successfully confined axonal growth within PDMS microstructures and confirmed unidirectional action potential propagation.

Conclusions:

• The developed PDMS stamp-transfer method effectively integrates microstructures with CMOS MEAs.
• This approach facilitates the study of neural networks with high spatial resolution and defined topology.
• Enables advanced investigation into neural information processing using engineered biological systems.