Interfacing in silico and in vitro neuronal networks

Area of Science:

• Computational Neuroscience
• Neuroscience
• Artificial Intelligence

Background:

• Investigating the interaction between biological neural networks (BNNs) and artificial spiking neural networks (SNNs) is crucial for advancing brain-computer interfaces.
• Previous research has explored unidirectional communication, but achieving bidirectional closed-loop systems remains a significant challenge.

Purpose of the Study:

• To establish an open-loop system connecting an in vitro cortical culture (BNN) with an in silico SNN.
• To evaluate the entrainment of BNN activity by an SNN.
• To lay the groundwork for future closed-loop bidirectional communication between biological and artificial neural networks.

Main Methods:

• An artificial spiking neural network (SNN) was developed to mirror the activity of a biological network (BNN).
• Network bursts detected in the SNN were used as triggers to stimulate the in vitro cortical culture (BNN) in an open-loop configuration.
• The correlation between SNN-triggered stimuli and BNN evoked network burst rates was analyzed across different timescales to assess entrainment.

Main Results:

• A significant correlation was observed between the SNN's activity and the BNN's evoked network burst rates.
• This correlation remained nearly constant across all analyzed timescales, indicating consistent entrainment.
• The magnitude of the entrainment was directly dependent on the efficacy of the stimulation source in influencing the BNN activity.

Conclusions:

• The study successfully demonstrated that an SNN can effectively entrain the activity of a BNN in an open-loop system.
• The findings support the feasibility of using SNNs to modulate biological neural activity.
• This research provides a foundational step towards developing more sophisticated closed-loop brain-computer interfaces.