A pattern grouping algorithm for analysis of spatiotemporal patterns in neuronal spike trains. 2. Application to...

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Analyzing complex neuronal activity is crucial for understanding brain function.
• Spatiotemporal patterns in neuronal spike trains may encode information.
• Previous methods lacked the capacity to analyze large, simultaneous spike train datasets.

Purpose of the Study:

• To demonstrate the practical application of the pattern grouping algorithm (PGA) for analyzing spatiotemporal patterns in neuronal spike trains.
• To investigate the relationship between neuronal firing patterns and brain states or behaviors.
• To explore the potential for complex temporal coding in higher brain centers.

Main Methods:

• Application of the pattern grouping algorithm (PGA) to simulated and experimental spike train data.
• Analysis of up to 30 simultaneously recorded spike trains.
• Experimental manipulation involving reversible inactivation of the cerebral cortex in anesthetized rats.
• Recording from the temporal cortex of freely moving rats during a discrimination task.

Main Results:

• The incidence of detected patterns did not correlate directly with increased firing rates after Hebbian learning in simulated networks.
• Specific spatiotemporal patterns in the thalamus reappeared upon restoration of cortical activity, suggesting widespread cell assemblies.
• The presence or absence of certain patterns before a cue in rats performing a task correlated with reaction time changes.
• Neuronal network interactions were shown to generate detectable spatiotemporal firing patterns.

Conclusions:

• Neuronal network interactions can generate complex spatiotemporal firing patterns.
• The pattern grouping algorithm (PGA) is effective for detecting these patterns in large datasets.
• Patterned neuronal activity is associated with specific behavioral states and suggests sophisticated temporal coding schemes in the brain.