Comparative computations of spike synchronization in visual cortex of cats

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• Neuronal synchronization is proposed as a mechanism for integrating visual information.
• Quantifying synchronization strength is crucial for understanding visual perception as an electrophysiological sign.
• Cross-correlograms are commonly used to analyze neuronal synchronization.

Purpose of the Study:

• To describe and compare six techniques for computing synchronization strength using the central peak of cross-correlograms.
• To evaluate the comparability and sensitivity of different synchronization measurement methods.
• To assess the reliability of visual inspection and regression analysis for comparing synchronization measures.

Main Methods:

• Electrophysiological recordings were performed in anesthetized cats.
• Six distinct computational methods for calculating cross-correlogram peak synchronization were applied.
• Statistical analyses, including regression, were used to compare the methods.

Main Results:

• Visual inspection of cross-correlograms can be misleading for comparing synchronization strength across different conditions.
• Regression analysis revealed poor correlation between methods, indicating they are not directly interchangeable.
• The sensitivity of each computational method varied, leading to potential under- or over-evaluation of functional connections.

Conclusions:

• Current methods for measuring neuronal synchronization strength are not universally comparable.
• The choice of computational strategy significantly impacts the evaluation of neural functional connections.
• Further research is needed to develop standardized and reliable methods for quantifying neuronal synchronization in visual perception.