Multiscale sample entropy analysis of resting-state fMRI over the lifespan

Motivation: Multiscale sample entropy (MSE) is a common complexity metric used in functional Magnetic Resonance Imaging (fMRI), yet it has not been applied to analyze the evolution across age groups. Goal(s): The goal of this study is to analyze the evolution of fMRI complexity over the lifespan using MSE. Approach: Resting-state fMRI data of 526 subjects age 6 – 85 years were analyzed with a linear mixed-effect model (LME) of MSE in different brain regions. Results: Overall, a decrease in mean gray matter complexity was observed after puberty. LME model showed significant decrease in complexity with age in middle frontal and superior frontal gyrus. Impact: The findings of this study shed light on the evolution of brain complexity with development and aging and may provide benchmark for detecting aberrant complexity in brain disorders.