A state observer for the computational network model of neural populations

Area of Science:

• Computational Neuroscience
• Systems Neuroscience
• Control Theory

Background:

• State observers are crucial for state feedback neuromodulation in treating neurological and psychiatric disorders.
• Existing methods lack observers for computational network models of neural populations.

Purpose of the Study:

• Design a state observer to reconstruct unmeasured states in computational network models of neural populations.
• Provide a theoretical basis for state feedback neuromodulation clinical schemes.
• Address the need for observers in computational neuroscience.

Main Methods:

• Utilized input-output stability theory and Lurie system theory.
• Solved linear matrix inequality conditions for observer design.
• Formed observer matrices from the optimal solution of linear matrix inequality conditions.

Main Results:

• Successfully designed a state observer for the computational network model of neural populations.
• Demonstrated the observer's ability to reproduce internal state variables in normal and lesion populations.
• Showcased observer robustness to input uncertainty and measurement noise.

Conclusions:

• This work presents the first observer design for computational network models of neural populations.
• The developed observer is a foundational step towards designing state feedback neuromodulation schemes.
• This research opens a new direction in computational neuroscience for clinical applications.