Minimum mean square error estimation of connectivity in biological neural networks

Area of Science:

• Computational Neuroscience
• Signal Processing

Background:

• Accurate identification of synaptic connectivity is crucial for understanding neural network function.
• Conventional correlation methods for synaptic connectivity estimation are data-intensive and computationally expensive.
• Existing methods struggle with nonstationary neuronal firing patterns.

Purpose of the Study:

• To develop a novel Minimum Mean Square Error (MMSE) estimation scheme for identifying synaptic connectivity in neural networks.
• To reduce data requirements and computational costs associated with conventional methods.
• To provide an estimation method suitable for both stationary and nonstationary neuronal firings.

Main Methods:

• Developed two recursive algorithms for estimating synaptic connectivities: one for nonlinear filtering and one for linear filtering.
• Determined the lower and upper bounds for the MMSE estimator.
• Analyzed the consistency of the estimators in quadratic mean.
• Investigated the relationship between the cross-interval histogram and MMSE estimation.

Main Results:

• The proposed MMSE estimation scheme significantly reduces data and computational costs.
• The method is effective for both stationary and nonstationary neuronal firing.
• The estimators were shown to be consistent in quadratic mean.
• The conventional cross-interval histogram was identified as an asymptotic linear MMSE estimator with an incorrect initial value.
• Simulations demonstrated that the MMSE estimators asymptotically approach true connectivity values for both linear and nonlinear cases.

Conclusions:

• The MMSE estimation scheme offers a more efficient and robust approach to determining synaptic connectivity in neural networks.
• This method advances the analysis of neural network dynamics, particularly for complex firing patterns.
• The findings provide a theoretical and practical improvement over existing connectivity estimation techniques.