Dynamic autocorrelation as a tool to measure time-dependent changes in neuronal time-scale along the hippocampus axis during resting-state.

Motivation: The autocorrelation (AC) of fMRI signals varies along the hippocampus (HPC), with higher values in the anterior-medial (AM) compared to the posterior-lateral (PL) regions. However, it remains unclear if this pattern is stable over time. Goal(s): We introduced dynamic autocorrelation (dAC) to investigate temporal changes in the AC pattern. We aimed to determine if the AC along the HPC changes over time. Approach: Voxel-wise dAC was estimated within the HPC for each subject, and the values were averaged across the PL and AM regions. Results: The dAC pattern along the HPC is not stable and fluctuates over time. Impact: We introduced dynamic autocorrelation (dAC) as a measure to investigate dynamic changes in AC patterns and demonstrated that the AC pattern is dynamic during resting-state. The dAC provides a promising approach to examine how neural timescales adapt to environmental changes.