Controlling neural network responsiveness: tradeoffs and constraints.

Area of Science:

• Neuroscience
• Systems Biology
• Computational Neuroscience

Background:

• Developing methods for whole-brain neural population control is crucial for advanced medical applications.
• Studying whole-brain dynamics presents significant challenges due to complexity and constraints.
• Understanding the controllability of neural networks is an ongoing area of research.

Purpose of the Study:

• To present an effective method for controlling response features (probability and latency) of cortical networks in vitro.
• To offer a novel experimental approach for studying the controllability of neural networks.
• To investigate the impact of closed-loop control on neural network dynamics.

Main Methods:

• Utilized in vitro cortical networks for experimental control.
• Employed closed-loop control strategies to manipulate neural activity.
• Monitored and analyzed response features like probability and latency over extended periods.
• Investigated alterations in global input-output relations and neuronal pair-wise correlations.

Main Results:

• Achieved effective control over response probability and latency in cortical networks in vitro for extended durations.
• Demonstrated that enforcing stable high activity rates via closed-loop control can alter global input-output relationships.
• Observed activity-dependent dispersion of neuronal pair-wise correlations across the network under controlled conditions.

Conclusions:

• The presented in vitro approach serves as a valuable experimental tool for studying neural network controllability.
• Closed-loop control of neural activity can modify fundamental network properties, including input-output relations and correlation structures.
• This work provides insights into the dynamics and control of neural populations, relevant for future medical applications.