Research on the lateralization of brain functional complexity in mild cognitive impairment-Alzheimer's disease progression based on multiscale lateralized brain entropy

Objective Despite extensive recent interest in mild cognitive impairment (MCI) and Alzheimer's disease (AD), few studies have examined the lateralization of their brain functions. This study describes the multiscale lat-eralized brain entropy (multi-LBE) algorithm and investigates the lateralization of brain functional complexity for MCI and AD.Methods The lateralized sample entropy (LSE) and lateralized cross-sample entropy (LCSE) are calculated over multiple time scales and used as lateralization indices of brain functional complexity to explore the progression of MCI-AD (from healthy control to early MCI, late MCI and AD).Results Our experimental results indicate that patients with AD have the weakest lateralization of functional complexity in most brain regions. The left hemisphere exhibits a trend to rightward laterality over an increasing number of time scales during the MCI-AD progression. Correlation analysis of LBEs demonstrates positive correlations among three patient groups, while negative correlations are found between healthy control group and each patient group. Furthermore, significance tests are performed on the whole brain regions of the four groups at multiple time scales, and the LBEs of several time scales are selected as features.Conclusion The classification of the four groups using multi-LBEs achieves an accuracy of up to 97.46% by multiscale LBEs, indicating that the LBE is a possible indicator of different states during MCI-AD progression. Significance Our findings provide new insights into the lateralization of brain functional complexity and highlight the potential for using multi-LBEs as biomarkers for the MCI-AD.