Functional parcellation of the hippocampus from resting-state dynamic functional connectivity

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Brain Imaging

Background:

• The hippocampus, crucial for learning and memory, has functionally heterogeneous regions.
• Previous functional subdivisions of the hippocampus often ignored the dynamic nature of resting-state functional connectivity (FC).

Purpose of the Study:

• To identify dynamic covariant structures within the hippocampus during spontaneous brain activity.
• To develop a more reproducible and functionally relevant parcellation of the hippocampus.

Main Methods:

• Utilized a two-stage spectral clustering technique on windowed FC correlations from Human Connectome Project data.
• Analyzed resting-state functional connectivity (FC) in 50 subjects with minimal head motion.
• Identified transient connectivity states reflecting hippocampal-cortical interactions.

Main Results:

• Subdivided the left and right hippocampus into six and five distinct functional subregions, respectively.
• Demonstrated significantly improved reproducibility of hippocampal subregions compared to static FC analysis.
• Highlighted the functional homogeneity of identified subregions across transient connectivity states.

Conclusions:

• Dynamic FC analysis provides a more accurate and reproducible method for hippocampal parcellation.
• This approach enhances our understanding of the hippocampus's functional organization and temporal flexibility.
• Findings offer a more comprehensive view of hippocampal function in learning and memory.