Bootstrap testing for cross-correlation under low firing activity

TL;DR

This paper introduces a bootstrap-based cross-correlation index to assess neural synchronization differences during sleep-like and awake-like brain states, especially under low firing activity conditions.

Contribution

It proposes a novel synchrony index based on kernel estimation and develops two bootstrap resampling methods for analyzing neural synchronization.

Findings

First bootstrap method detects significant differences in low firing rate states.

Second bootstrap method uncovers subtle synchronization differences in awake-like states.

The methods effectively differentiate neural synchronization levels across brain states.

Abstract

A new cross-correlation synchrony index for neural activity is proposed. The index is based on the integration of the kernel estimation of the cross-correlation function. It is used to test for the dynamic synchronization levels of spontaneous neural activity under two induced brain states: sleep-like and awake-like. Two bootstrap resampling plans are proposed to approximate the distribution of the test statistics. The results of the first bootstrap method indicate that it is useful to discern significant differences in the synchronization dynamics of brain states characterized by a neural activity with low firing rate. The second bootstrap method is useful to unveil subtle differences in the synchronization levels of the awake-like state, depending on the activation pathway.