Adaptation modulates effective connectivity and network stability.

Area of Science:

• Computational neuroscience
• Systems neuroscience
• Neural dynamics

Background:

• The brain operates as a complex, nonlinear network near a critical state known as the edge of chaos for optimal function.
• Precise balance between neural excitation and inhibition is crucial for stability and rich dynamics.
• Existing models struggle to reconcile network stability with computationally useful low-dimensional structure in synaptic matrices.

Purpose of the Study:

• To investigate dynamic mechanisms that maintain brain stability near the edge of chaos.
• To propose spike frequency adaptation (SFA) and short-term synaptic depression (STD) as key regulators of neural network dynamics.
• To explain the need for linear time-varying (LTV) models in analyzing brain signals and suggest a link to compromised adaptation in neurological conditions.

Main Methods:

• Theoretical modeling of neural network dynamics.
• Analysis of synaptic weight matrices and their structural properties.
• Framework linking intrinsic and synaptic negative feedback to network stability.

Main Results:

• Biologically realistic synaptic structures pose challenges for static balancing rules that preserve stability and low dimensionality.
• External stimuli can dynamically unbalance neural networks.
• SFA and STD are proposed as dynamic mechanisms that modulate effective connectivity, maintaining stability near the edge of chaos.

Conclusions:

• The brain employs dynamic adaptation mechanisms (SFA, STD) to actively regulate stability, rather than relying on static balancing.
• Compromised adaptation may contribute to neurological disorders characterized by altered excitability.
• Targeted brain stimulation could be a tool to study adaptation's regulatory role.