Designing optimal perturbation inputs for system identification in neuroscience.

Optimally designed perturbation inputs improve neural system identification accuracy by substantially outperforming conventional, intuition-based inputs in simulations.

• Why it matters: Understanding neural network dynamics is limited by passive activity, which provides insufficient information for estimating connectivity, making effective perturbation design crucial for accurate modeling.
• What they did: A data-driven, control-system-based framework was developed to optimize perturbation inputs, minimizing estimation errors and exploring their relationship with common stimulation patterns through simulations.
• The result: The framework enables more reliable discrimination of neural states and precise control of neural transitions, advancing the potential for diagnosing and treating neurological conditions.